ORNL-0768.txt

 

 

 

OAK RIDGE

o e B TS T Y e Ty e — ]

]‘ srssrioaron paCARBIDE]

BLCLASSIEICATION OFFICER
OAK RIVGE RATIONAL LABCRATGRY
AUTHORITY DELEGATED BY EMDA 8157

’_ T B W, Ll T e o W Nl Bl 1]

 

 

 

AT THE OAK RIDGE NATIONAL LABORATORY

u
DATE'(“JL X 1979C~DU-‘= ?{cw 7

 

ORNL 768

Lo
g

Ilalctoru
t
ST %"‘*““% copy;;
3 4456 05bL28L 9 W o

      

THE AIRCRAFT NUCLEAR
PROPULSION PROJECT

NATIONAL LABORATORY
OPERATED BY

§ND CARBON CHEMICALS DIVISION
l'f N CARBIDE AND CANUGN CORPORATION

POBY OFFICE BOX P
OAK RIDOE, TENNERBEE

 
  
 
 

 

 

P T
g‘-}'i" ovET “‘_‘% By #ut*u‘w*v O, ORNL 768
¥ Heo el 5 Reactors
5”"3{ Wl hww fln i%nd //L L{ J 2574 7 Progress Report
— - Q:%/7—ZL7"“" iesmemsmms This document consists of 103
For: H. T. Brav. Someoier pages
Laboratory ecasds Bigd Copy Z %] of 161 copies. Series &

Contract No. W-7405, eng. 26

THE AIRCRAFT NUCLEAR PROPULSION PROJECT

at the
0ak Ridge National Laboratory

A. M. Weinberg, Project Director
C. B. Ellis, Project Coordinator

QUARTERLY PROGRESS REPORT
for Period Ending May 31, 1950

Edited by
M. J. Nielsen and C. B. Ellis

pate 1ssvep QUG 1 41950

OAK RIDGE NATIONAL LABORATORY
operated by
CARBIDE AND CARBON CHEMICALS DIVISION
Union Carbide and Carbon Corporation
Post 0ffice Box P
0ak Ridge, Tennessee

 
el

ORNL 768
Reactors .
Progress Report

INTERNAL DISTRIBUTION .
G. T. Felbeck (C&CCC) 21. F. C. VonderLage 42. W. R. Gall

1.

2. 706-A Library 22. A. Hollaender 43, R. M. Jones

3. 706-A Library 23. J. A. Swartout 44, E. C. Miller

4. 706-B Library 24. J. H. Gillette 45. D. S. Billington
5. Biology Library 25. K. Z. Morgan 46. E. P. Blizard

6. Health Physics Library 26. J. S. Felton ~ 47. W. M. Breazeale

7. Training School Library 27. A. H. Snell 48. W. K. Ergen

8. Training School Library 28. F. L. Steahly 49, C. E. Clifford

9. Metallurgy Library 29. M. T. Kelley 50. M. L. Nelson

10. Central Files 30. C. E. Winters 51. E. D. Shipley

11. Central Files 31. J. A. Lane 52. G. H. Clewett

12. Central Files 32. G. E. Boyd 53. C. P. Keim

13. Central Files 33. R. W. Stoughton 54. 0. Sisman

14. C. E. Center 34-35. F. L. Culler 55. A. D. Callihan

15. C. E. Larson 36. D. G. Reid 56. R. S. Livingston
16. W. B. Humes (K-25) 37. A. S. Householder 57. C. D. Susano

17. W. D. Lavers (Y-12) 38. M. M. Mann 58. M. J. Nielsen

18. A. M. Weinberg 39. C. B. Graham 59. D. W. Cardwell

19. E. J. Murphy 40. R. E. Engberg 60-69. C. B. Ellis (Y-12)
20. C. B. Ellis 41. R. N. Lyon 70-71. Central Files (0.P.)

EXTERNAL DISTRIBUTION

72-76. Air Force Engineering Office, Oak Ridge
77-88. Argonne National Laboratory
89-90. Atomic Energy Commission, Washington
91. Battelle Memorial Institute
92, Bureau of Aeronautics
93-96., Brookhaven National Laboratory
97. Bureau of Ships
.98. Chicago Operations Office
99-102. General Electric Company, Richland
103. Hanford Operations Office
104-105. Tdaho Operations Office
106. Towa State College
107-110. Knolls Atomic Power Laboratory
111-113. Los Alamos
114. Massachusetts Institute of Technology
115-116. National Advisory Committee for Aeronautics
117-134. NEPA Project
135. National.Advisory Committee for Aeronautics, Washington
136-137. New York Operations Office
138. North American Aviation, Inc.
139. Office of Naval Research
140. Patent Branch, Washington
141-155. Technical Information Branch, ORE
156-157. University of California Radiation Laboratory
158-161. Westinghouse Electric Corporation

2

-
TABLE OF CONTENTS

SUMMARY
INTRODUCTION

REACTOR DESIGN

Multi-group Criticality Calculations
Reactivity Calculation Techniques
Gamma Activity of Primary Coolants

Nuclear gamma emitters
Bremsstrahlung. from lithium beta rays

Radiocactivity in the External Coolant of the Circulating Fuel
Reactor

Evolution of Fission Gases from a Fuel Element
Advanced Seminar in Reactor Physics

Coolant Dynamics

Liquid Metal Cooled ARE Designs

NaOH Cooled ARE Design

General Coolant Cycle Studies

.CRITICAL EXPERIMENTS

-SHIELDING

Bulk Shielding Experiments
Interpretation of Lid Tank Data
‘New Bulk Shield Testing Facility
Shielding Analysis

Neutron Energy Spectrometer
Duct Theory

Shielding Survey

Shielding Materials

HEAT TRANSFER
EXPERIMENTAL ENGINEERING

METALLURGY AND MATERTALS

Static Sorting Tests

Pure metals in 1000°C lithium
Pure metals in 1000°C bismuth
| Dissolved fuel tests
Dynamic Corrosion Tests

Thermal convection loops
Forced circulation system

11

14

14
14
14

15
18

18
21
23
24
26

27
29

33

36

36
43
48
52
54
54
55
56

57
61

62
62

67
68
17

79

79
80
TABLE OF CONTENTS (Con’ t)

LITHIUM ISOTOPE SEPARATION

{Molecalar Distillation
' Chemical Exchange in a Liquid-Liquid System

 

i Lithium amalgam-lithium salt solution
| Aqueous-organic systems
Organic-organic systems
Pulse column application

Chemical Exchange in a Liquid-Solid System
Auxiliary Studies

RADIATION DAMAGE
Reactor Materials

Accelerator experiments
In-pile creep tests
Diffusion of fission products

Auxiliary Materials
Plastics
Metal hydrides
"Liquid metals in-pile experiment
NUCLEAR MEASUREMENTS
High Voltage Program

Mechanical Velocity Selector

PUBLICATIONS

81

81
82

87
88
88
89

93
93

95
95

95
96
98

99

99
99
99

102

102
102

103
LIST .OF TABLES

Gamma-Emitting Contaminants-Binary System
Gamma Emitting Contaminants —Circulating Fuel System
‘Properties of Some Potential Reactor Coolants

Corrosion of Materials by Liquid Lithium in Static
Tests - 4 hrs at 1000°C

Corrosion of Materials by Bismuth in Static Tests - 4 hrs at

1000°C

16

22

25

63

64
i o> W B

N

10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

LIST OF FIGURES

The Aircraft Nuclear Propulsion Project at Oak Ridge National
Laboratory

Elements of Plate-Type Reactors
Critical Experiments Facility
Neutron Data 27% Pb, 73% H,0, Centerline Measurements

Gamma Center Line Measurements 27% Pb, 73% H,0, Shield
Mock-Up

Neutron Data 18% Pb, 72% HQQ,_Centerline Measurements

Gamma Centerline Measurements 18% Lead, 82% Water, Shield
Mock-Up '

Gamma Centerline Measurements Fe, H,0, and Pb Mock-Up

Neutron Centerline Measurements, Iron and Water Shield

Mock-Up ‘

‘Bulk Shielding Facility
‘Bulk Shielding Facility

.Proposed Shielding Facility, Isometric Layout

Figure "8" Liquid Metals System

Liquid Metals Pump

.Liquid Metal Corrosion Test Capsule
Corrosion Samples Tested in Molten Lithium
Corrosion Samples Tested in Molten Lithium

Corrosion Samples Tested in Molten Lithium

Corrosion Samples Tested in Molten Lithium

Corrosion Samples Tested in Molten Bismuth
.Corrosion Samples Tested in Molten Bismuth

Corrosion Sampies Tested in Molten Bismuth

Sectional View Vacuum Argon Corrosion Testing Apparatus
¥

‘Single Stage Lithium Still

Two Stage Lithium Still

.Two Stage Lithium Still

Pulse Column

Pulse Column %

i’in-Pile Creep Apparatus for Insertion in Stringer in ORNL Pile

Liquid Metals "In-Pile" Experiment
6

13
28
34
37

.38
-39

f

40

41

42
49
50
51
58
60
66
69
70
71
72
74
75
16
78
83
85
86
90
91
97
100
SUMMARY -‘

Reactor Design. Operational methods are being developed for treating

proposed intermediate reactor designs by multi-group analysis.

A new survey has been made of the impurities to be allowed in a primary
coolant stream emerging from a reactor shield. Among the impurities most
difficult to shield would be Na, V, Mn, Rh, and In. Calculations on the
Bremsstrahlung which would arise in piping containing radioactive Li indicate
that a primary Li’ stream could probably not be brought out of the reactor

without some external shielding.

Some calculations have been made on radioactivityin the secondary coolant
of acirculating fuel reactor system which might arise from the delayed neutron
flux in the intermediate heat exchanger. It appears that this effect would be

negligible.

A survey has been made of the pressure drops to be expected across re-
presentative reactor core geometries with various liquid metal coolants. The
superiority of the light metals Li and Na over Pb and Bi from this standpoint

1s very great.

A number of preliminary designs for a 1000 kw ARE reactor have been
sketched out in first approximation. Most of these emphasize liquid metal
coolants with solid fuel elements; however, one proposal involves sodium
hydroxide cooling, with the eventual aim of developing a homogeneous circulat-
ing fuel aircraft reactor. .All of these designs involve a considerable number
of materials unknowns. Research aimed at clarifying these materials problems

is now underway.

Critical Experiments. A joint ORNL-NEPA group is being established to
carry out critical experiments of interest to the ANP program. A new building
to be occupied by this group is expected to reach completion in July, 1950.
Most of the apparatus for beginning the critical experiments has already been
built and tested by NEFA.

‘Shielding. Data are presented in Figs. 4, 5, 6 and 7 for the neutron and
gamma attenuation of shield assemblies containing 27% and 18% Pb by volume in
water. Measurements of 33% Pb inwater have been started. Data on the Fe-H,0-Pb

shield for the Naval reactor are presented in Figs. 8 and 9. As a result of a
recalibration and improvements in the instrumental setup, it is believed that

the accuracy of the intensity measurements is now within 30%.

A theoretical method called the neutron accountability system has been
developed for estimating fast neutron leakage through a sample from the purely
thermal measurements in the lid tank. Such calculations, when added to the
new lid tank data, indicate that a unit shield to surround a spherical four
foot aircraft reactor, in the idealized case neglecting ducts and heat ex-

changers, would weigh in the neighborhood of 140,000 pounds.

Construction has begun on the building to house the new shielding reactor.
A reactor assembly and measurement equipment are expected to be ready for in-
stallation when the building is completed in the fall of 1950. It is believed
that the neutron and gamma intensity through shielding samples will be high
enough with this new facility to permit making spectral intensity measurements.
A neutron camera similar to a Los Alamos model is now being built, and several
other instruments suitable for neutron or gamma ray spectroscopy are under de-

sign.

Theoretical calculations have been made on the spectrum to be expected of

neutrons emerging from various thicknesses of H,0 and Fe-H,0 shields.

A study has been made on the desirability of using W for a major part of
the shielding. It does not appear that W is likely to be enough better than
Pb to be worth the difficulty and expense of fabricating it in large quantities.
However, UH, still appears theoretically to offer worthwhile shield weight
savings. The possibility of obtaining this material at high density and in a

non-pyrophoric state is being investigated.

The NEPA fast neutron spectrometer has been tested on the electrostatic
generator at Bartol and at MIT. 1Its resolution appears adequate for shielding

measurements.

Theoretical analyses are being carried out on several air duct shielding

problems of simple geometry.

Boral has now been produced in large sheets of1/8 in. thickness. Methods

of joining Boral sheets are being investigated.

Heat Transfer. A figure-of-eight loop for circulating liquid metals to
measure heat transfer coefficients is now nearing completion. This circulating

system will employ a hydraulic bearing pump which eliminates any seals.
 

Experimental Engineering. A new group is being established to carry out
all varieties of large-scale experimental engineering, particularly in liquid

metals problems preliminary to the construction of the ARE reactor.

Metallurgy and Materials, Four-hour static corrosion tests of numerous
materials in liquid Li and Bi at 1800°F have now been completed. The detailed
effects upon the sample surfaces are discussed in this report. Those materials
which so far show the greatest promise in liquid Bi are Fe, Mo, W and Ta. For
Li at 1800°F the same materials, together with Zr and a number of stainless

steels, seem most promising.

Studies are underway on mass transfer effects in liquid metal systems

whose containing walls involve several different materials.

Four-hour static corrosion tests at 1800°F have also been completed on a
number of materials in liquid Bi containing dissolved uranium-—a fluid which
might be used in a circulating fuel reactor. Mo and W show the least attack

so far.

Twelve convection loops (harps) are now being made of various materials
for dynamic corrosion tests. A small forced circulation system for corrosion

tests is also being planned.

Because of evidence that much of the observed corrosion in molten Li
systems 1s caused by oxygen and nitrogen impurities, a special high tempera-

ture L1 purifier i1s now being constructed.

Lithium Isotope Separation. A separation factorof1.02 has been obtalned
?by molecular distillation. A two-stage refluxing still is being built. A
%packed column refluxing still to test feasibility of higher temperature and

-higher pressure distillation has been constructed.

More than 400 organic-organic systems have been investigated in a search
_for asuitable liquid-liquid countercurrent arrangement for isotope separation.
Of these, 12 show sufficient promise for further investigation. In addition,

a number of aqueous-organic systems are being studied.
Pulse columns are being studied for possible use withaLi amalgam system.

Additional research on Li isotope separation isunderway with ion-exchange

resins.

Radiation Damage. Deuteron bombardment of Fe immersed in Li at 1800°F 1is

being carried out by North American Aviation, Inc. at Berkeley to study effects
of radiation on corrosion rates. Similar cyclotron bombardment will be started
shortly at Purdue to study the effect of radiation on creep rate. Silicon
carbide or magnesium silicide will probably be used there, since these are un-
classified materials analogous to the Be,C which is of interest for the air-

cooled reactor.

The creep test equipment to go into the ORNL reactor is partially com-

pleted.

Samples of Ti, Mo, Ni and Fe for before-and-after mechanical property

measurements are expected to go into the Hanford pile shortly.

An apparatus has been designed for exposing plastics to a 10° R/hr gamma

source.

Hydrides of Ti, Zr, and Li have been exposed in the OBNL reactor and will

be analyzed for radiation damage after cooling.

An experimental liquid metal loop at 1800°F is being constructed for in-
sertion within the OBNL reactor. It will be used initially in exploring the
random radioactivity arising in liquid metal coolants. Later it should also

serve for studies on radiation damage.

Nuclear Measurements. The high voltage laboratory building to house the

2 Mev and 5 Mev Van de Graff machines has been designed.

A mechanical neutron velocity selector is now being designed which will
incorporate a 12-inch steel rotor running at 10,000 rpm. An 80-channel re-

cording system 1s planned.

10
INTRODUCTION

The Aircraft Nuclear Propulsion Project (ANP) is a joint effort of three
laboratories: The QOak Ridge National Laboratory, the NEPA Division of the
Fairchild Engine and Airplane Corporation, and the Lewis Laboratory at Cleveland
of the National Advisory Committee for Aeronautics (NACA). The work of these
three groups is guided in broad outline by the ANP Policy Committee in Washing-

ton, whose members are:

Director of Reactor Development, Atomic Energy Commission, Chairman
Director of Research and Development, United States Air Force
Director of Research, National Advisory Committee for Aeronautics

Deputy Chief, Bureau of Aeronautics, Navy Department

This Policy Committee acts upon recommendations submitted by the ANP Technical
Committee, which is composed of the Technical Directors of the three labora-
tories. During the past quarter these committees have held several meetings,
with the result that the division of effort among the laboratories for fiscal

year 1951 has been fairly clearly delineated.*

The Oak Ridge National Laboratory has general responsibility for reactor
research.” Within this general field the Laboratory is responsible for overall

coordination throughout the Project of the work on:

Shielding theory and tests,
Nuclear measurements,
Critical experiments,
Radiation damage, and

Fuel reprocessing.-

In reactor materials research each. laboratory will proceed somewhat independ-

ently, but with continuous coordination by a joint materials committee.:

The Aircraft Reactor Experiment. In addition to these general efforts,
the Oak Ridge National Laboratory has proposed the construction of a small re-
actor at Oak Ridge. This project is called the Aircraft Reactor Experiment
(ARE).j The purpose of this experimental reactor, to operate at a power level
of 1000 kw, is to obtain data and experience necessary for the design of a

full-scale aircraft reactor. It is intended that the ARE will incorporate all

# The most recent survey of the work of the NEPA Division is NEPA-1374-IPR-52, Quarterly Report for
the Period January 1 to March 31, 1950.

11
of the features which appear most promising for a large aircraft reactor of
the liquid-cooled type. A feasibility report will be submitted for approval
during the next quarter. It is planned that ORNL will carry the full respon-
sibility for design and construction of the ARE, but that assistance will be

received from NEPA and NACA in many phases of the work.

The development of the ARE reactor is being approached by attempting to
first visualize an approximate design for a full-scale aircraft reactor. This
design will then be arranged to operate at 1000 kw with as little alteration
as possible.j Liquid metal cooling is favored at present for ARE, although
alternate proposals are still receiving consideration.: A very considerable
fraction of the experimental ANP work of the Laboratory, as described in the
present report, has already been pointed along lines to answer problems con-

fronting the ARE Design Gfoup.

Personnel. Oak Bidge National Laboratory personnel participating in the
ANP program now total 125. Of these, 87 are in efigineering and scientific
categories. In addition there are 19 NEPA men working full or part-time at
the Laboratory, together with five men from the United States Air Force, one
from NACA, one on leave from the Georgia Institute of Technology, and one on
leave from Reaction Motors, Inc. Figure 1 is an organization chart for the

ANP Project at this laboratory.

There are now six outside contracting groups associated with the ANP
effort of the Laboratory: Nuclear Development Associates, Inc., H. K. Ferguson
Company, and North American Aviation, Inc., for general design and analysis;
Purdue University for radiation damage work; and the University of California
and Stanford University for fundamental heat transfer measurements. Summaries

of the work of these contractors are included in this report.
¢

12
 

THE AIRCRAFT NUCLEAR PROPULSION PROJECT
AT THE OAK RIDGE NATIONAL LABORATORY

PROJECT DIRECTOR
A. M. Weinberg
E. B. Richardson, Sec.

 

 

 

ANP TECHNICAL
ADYISORY BOARD
F. W. Loomis, Chairman
D. Hilyer, Sec.

 

 

PROJECT COORDINATOR
C. B. Ellis
L. ¥. Bond, Sec.

 

 

 

 

 

¥. E. Thompson, Project Becretary EXECUTIVE ASSISTANT

P. ¥. Rueff
M. Pickard, See.

 

H. F. Mebuffie, Project Editor

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SHIELDING | %RADIATION DAMAGE | GENERAL DESIGN METALLURGY AND HEAT TRANSFER JOINT ORNL=-NEPA EXPERIMENTAL SPEGIAL
f ; GROUP MATERIALS (at ¥-12) CRITICAL ENGINEERING ISOTOPES
| (At ¥-12) At Xo10) EXPERIMENTS e 710,
SHIELD RESEARCH ! REACTOR CORE C. B. Flli MATERIALS o PHYSICAL CHEMISTRY
(At X-10) } MATERIALS Ellis, USAF IA R. N. Lyon (At new site in July) SEPARATIONS
' (At X-10) N. M. Smith E. C. Miller H. F. Poppendick H. W. Savage (At ¥-12)
E. P. Blizard ' M. C. Ediund* R. B. Day T R ey A. D. Calliham.* K-25
V. Anrews, Sec. B. S. Billimgton R. C. Briant F. ¥. Drosten W B. Harrison b Conerbens
W. W, Parkinson A. P. Fraas E. S. Bomar sobe ne S?June) J. F. Coneybear, *NEPA G. H. Clewett*
LID TANK C. D. Baumann J. H. Haines* L. G. Glasgow H. C. Claiborne F. W. Pressey, NEPA A. Clark
C. E. Clifford J. T. Howe Lt. €al. P. Hill, ¥. H. Bridges C. P. Coughlen F. T. Bly, NEPA J. Drury
Maj. L. H. Ballweg, D. K. Stevens USAF G. P. Smith (Aug.) Caot. A. R. Frithsen E. V. Haaki, NEPA F. Waldrop
USAF J. C. Wilzom R. W. Schreeder P. Patriarcha (July) Pl A T USAF' J. A. Hunter, NEPA W. Ward
T. V. Blosser R. M. Carroll J.=H. Wyld, RMI L. A, Abral.ns (July) W. S. Farmer (June) L. Fortenberry
J. D. Flynn R. Kernohan,* NEPA W. P. Berggren J. E. Cunningham ’ C. C. Haws, Jr.
H. E. Hunger ford J. Thomas,® Sec. D. K. Holmes . G. M. Adamson ¥. M. Leaders
R. Lewis, NEPA Le. Col. M.J. Nleé;:; TECHNIC IANS TECHNICIANS G. Marrow
TECENICT J. Ramsey®*
R. Burn?f: ¢ :fl.wflul[ings CONSULTANTS Ruth Willtams, Sec. J. Flynn L. Trotter M. Richardson R. Winn L p TWY_' hell
V. DiRite  DP. Kirby On Loan: R. 5. Webb C. L. Smith s Gatton, Sec. . P. iche
I Mubbara ¥ bartin | o r gl B Sebleer F. Gillime Sec
R. Cleland A, Forbes. Comn. J. E. Pape (July) MATHEMATICS
NEW FACILITY N. J. Grant Draftswan J. Thomas,* Sec. (At X-10)
W. ®. Breaz H. G. Macpherson (At X-10)
. eale, GIT CONSULTANT CONSULTANTS
J. E. Owens* M. L. Nelson* E. G. Bohlmann*
J. L. Meem, NACA CONTRACTORS G. F. Wislicenus-. R. o b Resson J. H. Gross
E. B. Johnson MECHANICAL TESTING b Doarts D. W. Whitcombe ,NEPA R. E. Wacker
T. E. Cole* CONTRACTORS ’ G. E. Boyd
3P Gill North American A. G. H. Andersen® y
W. R. Gall* Aviation, Inc. H. K. Ferguson €o.
R. Bullard, Sec. Chauncey Starr K. Cohen TECHNICIANS CONTRACTORS CONSULTANT
ANALYSLS and Others ¥. I. Thompson .
P. Miller R. T. Clen® = C. ¥. Weaver* S. L. Madorski
- d - -
W. K. Ergen, NEPA Purdue University and Others fl, L,_.:i‘_’.s&c_ Stanford University JOINT ORNL- NEPA
S. Podgor, NEPA K. Lark-Horovitz North Awerican 0. Sheppard CGONTROL GROUP
F. H. Murray and Others eiation o CONSULTANTS and Othons NUGLEAR ELECTROMAGNETIC
K. E. Keyes ' ! MEASUREMENTS SEPARA
Chauncey Starr (Now on MTR Control) RATIONS
L. Gordon, Sec. Universit £
A. 8, Thompson E. Creutz r y o A
SPECIAL INSTRUMENTS W. C. Parrish N. J. Grant Califernia (At X-10) W. H. Jordan*® (At Y-12)
J. L. Gre E. E. St. John, NEPA
B. R. Gossick H. Schwartz &g ) » .
F. J. Muckenthaler, (At ¥-12) and Others and 0th o R. G. Hester, NEPA and Others equiva-
T NEPA nd Others G, Pawlicki (At Y-12) lent to ten men
0. Sisman Nuclear Development tI- full time.
CONSULTANTS C. D. Bopp Associates, Inc. E. Bettis,* NEPA
. C. G. Collins* J: Menke E. R. Mann,* NEPA
F. Friedman .
H. Jordan TECHNICIANS o ounE
w. H. et G. Goertzel
]Iai. (1\)1 ollan . Ki(r;klsnd lln:S Towns A, R. Gruber
. ewson . Darneti, Sec.
and Others *Part Time

SHIELDING MATERIALS
(At Y-12)

A. S. Kitzes

V. L. McKinney

TECHNICIANS
K. Hullings ¥. Jones, Sec.®

 

 

 

 

 

 

 

 

 

Revised June 19, 1950, C. B. E.

FIG I
FAGE RS

 
REACTOR DESIGN

The ANP General Design Group and its contractors have been engaged in a
number of miscellaneous studies in reactor physics pointing towards the solu-
tion of problems arising in the ARE design. In addition, a number of problems
in reactor engineering have been analyzed and several possible designs for the
ABE have heen roughed oufi.j A number of these efforts are described briefly

below."

MULTI-GROUP CRITICALITY CALCULATIONS

M. C. Edlund, N. M. Smith,
J. W. Webster*

An operational method for the evaluation of the so-called constants em-
ployed in multi-group calculations has been formulated. It is pointed out that
the constants are functions of the macroscopic characteristics of the reactor—
that is, size, shape, etc.—and that they are also functions of position
within the reactor. The numerical integration of the multi-group equations
for specific reactor designs proposed for the ARE (with or without hydrogen
present) is contemplated by the method of Masket and Goertzel (MonP-360)."

REACTIVITY CALCULATION TECHNIQUES

Nuclear Development Associates, Inc.

Nuclear Development Associates plans to undertake a study of the various
methods for reactivity calculations applicable to reactors of aircraft type.
This work will include a survey of the older literature, and it may culminate
in the production of a Reactivity Handbook. Following the survey, numerous

reactivity calculations of direct interest to the ANP design will be undertaken.

GAMMA ACTIVITY OF PRIMARY COOLANTS

The question of whether the nuclear aircraft can be run on a binary cycle—

one in which a working fluid heated by the chain reaction is brought outside

* NEPA
14
and carried directly to the radiators to heat the turbojet air—depends to a
large extent on whether this working fluid will become too radioactive. Too
much gamma activity in the primary fluid would necessitate extensive shielding
of all the pipe lines and radiators, which would probably add a prohibitive
weight to the plane. There are three principal ways in which such activity
might arise: (a) release of fission products into the coolant stream, either
by diffusion through the fuel element cladding or by rupture of an element,
(b) activity induced in the coolant (or in its impurities) by the neutron flux,
or (c) activity induced in the container materials (or minute impurities in
these materials) anywhere in the coolant system—which might be dissolved by
the coolant, or ejected from the walls by neutron recoil, and so carried out-
side the shield.

It is generally assumed at the present time that in practice one or more
of these effects will arise and that the nuclear aircraft will have to employ
a ternary system, with an intermediate heat exchanger embedded in the shield
to transfer the heat from the primary fluid to a secondary external fluid
having no contact with the region of high neutron flux. However, this question
is not yet completely settled. Some calculations are presented here which
bear on two facets of the problem, the nuclear gamma activities to be expected
from various contaminants in the coolant stream, and the Bremsstrahlung gamma
rays which would arise in the external piping from the Li® beta particles, if

lithium were used as the primary coolant.

Nuclear Gamma Emitters (N. M. Smith). The results of the calculations
are displayed in Table 1. The elements giving rise to high Y-activity are:
Na, V, Mn, Rh, In, most of the raré:earths, W, Re, Ir and Au. To a lesser
extent Al, Cl, Sc, Cs and Ta are bad. Normal contamination amounts of Mg and

Fe may be allowed.

In this work the absolute density of each likely constituent in the re-
actor coolant was computed which gives rise to an exposure of one Roentgen per
day at the crew’s position 100 feet away, when the lines and radiators are

unshielded. The following assumptions were made:

1 Distance, emitter to crew 100 ft

2. Average flux level in reactor 101* neut/cm? sec
3. Dosage at crew’s position 1 Roentgen/day

4. Self-shielding factor of emitter 3

15
TABLE 1

Gamma-Emitting Contaminants-Binary System

(Permissible 7y-ray emitting contaminants 1in primary reactor coolant assuming
¢=10'* neutrons/cm® sec, S= 3, D x 100 ft, V, ~ 14 X 10°% cm®, ¥, 31 x10° cn’)

 

MAXIMUM PERMISSIBLE VALUES FOR ONE <y/DAY AT COCKPIT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

RELATIVE -ATOMIC
ABUNDANCE CONCENTRATION DENSITY: TOTAL MASS
ELEMENT PERCENT O, (barns) B () (Mev) T% No./cm3 (gm/cm3) (gm)
+nil5s =5 k 23 - 8
N 0.38 2 x 10 6 (75%) 7.35 s 2.5 x 10 .9 8.7 x 10
"Na?3 100 0.5 2.75, 1.39 14.8 h 3.6 x 1016 .4 % 1076 20
Mg2® 11.3 4.8 x 1072 0.84, 1.01(20%)| 10.2 m 1.3 x 101'° .6 x 1074 7900
*A127 100 0.18 1.8 2.4 m 2.3 x 1015 .0 x 105 140
+3i3° 3.05 0.12 No 7y 2.8 h - - -
1.6 (36%)
*C137 25 0.6 2.1 (47%) 37.5 m 2.6 x 107 5 x 10°° 212
3.6 (75%)
tK41 6.7 1.0 2.1 (25%) 12.44 h 3.5 x 107 .4 x 1075 340
1.12 (98%) - A
Sc*s 100 22 0.88 85 d 1.7 x 103 .3 x 1077 1.8
*Ca*® g cqa// 0;3///, 2°5§// 2.0 X 10i;//
.185 .055 - - 30m - .15 2.1 x 10°
Vel 100 5 1.45 3.74 m 1.0 x 10'° 51077 11.9
*Mn5 5 100 12.8 0.84 2.6 h 7.0 x 1015 6 x 10”7 9.2
1.10 (50%)
*FeS8 0.32 0.31 1.30 (50%) 46 d 5.8 x 1018 4 x 10°* 76009
04, .18
Rh103 100 149 0.95 42 s 4.3 x 1014 3 x 1078 1.02
4, 1.12
Inlls 95.8 56 1.31, 2.32 53.9 m 2.6 x 1014 9 x 10”8 0.69

 

 

 

 

16

 

 

 
TABLE 1 (Cont’ d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MAXIMUM PERMISSIBLE VALUES FOR ONE ¥/DAY AT COCKPIT
RELATIVE" ATOMIC
ABUNDANCE CONCENTRATION DENSITY TOTAL MASS

ELEMENT PERCENT O, (barns) E (Y) (Mev) TY% No. /cm3 (gm/cmd) (gm)
.60, .79

Cs!33 100 25.6 (95%) 2.3 y 2.1 x 1015 .6 x 1077 6.4

Lal3?® 99.9 9 1.63 (75%) 40 h 5.2 x 108 .2 x 1076 17.0
SE =

pri4! 100 11 2.65 19.3 h 3.3 x 1015 7 x 1077 10.8
1.03

Smls2 26.6 280 .07 47 h 1  x 1015 5 x 10°7 3.5
0.12, 0.16

Euls! 47.7 1530 0.72 9.2 h 1 x 10'* 5 x 10°8 0.35
0.37

Dyl64 27.3 2700 0.83 2.4 h 8.5 x 1013 3 x 10”8 0.32

Ho' ®% 100 67 1.2 27.2 h 9.3 x 10'* 5x 1077 3.5

Ybl74 97.5 30 0.35 4.1 d 7 5 x 1013 2 x 10°° 31
.112, .206 ,

Lul7?8 2.5 3200 .317 6.6 d 1.5 x 1015 4 x 1077 6.2
1.13 (37%)

Tal8! 100 20.6 1.22 (55%) 113 d 3.6 x 1015 1 x 10°° 15
SE =

w186 28.4 37.2 3.07 24.1 h 2.3 x 1015 .1 x 1077 10
SE =

Rel87 62.9 75 3.65 18 h 4.4 x 1014 .4 x 10°8 2.0
Oo(V 1.,5V /

Irt9!? 38.5 260 S =92.20 70d 3.4 x 104 .1 x 1077 1.5

Au'®’ 100 96 .41 2.69 d 1.9 x 10'° .2 x 1077 8.7

 

 

 

 

 

 

 

+ Normal Impurities in Li

Feld, in Lex P 157, estimates total mass of 3 x 10% gm of Fe permitted in (Bi) coolant.
(1.8 x 106)9 self -shielding (zero) and dosage (109-2 Mev ¥/r) to those used here one gets satisfactory agreement.

(TThis is equivalent to 2.8 x 10’4%‘in Li.

NOTE:

 

* Normal Impurities in Fe

17

Compare this with typical analysis.

Converting this assumption as to flux (2.4 x

.. . .7 .
The limitation on Li (and other B-emitters) shall be treated separately. The above table is limited entirely to y-ray emitters®

14

107 "), volume
5. Volume of coolant external to shield 14 X 10° cm®

6. Volume of coolant inside shield 1 % 10° cm®

It may be noted that none of the pure elements listed in Table 1, with
the possible exception of Ca, could be permitted to penetrate the core in
massive amounts and then become exposed to the unprotected crew. Pure sodium,
for instance, would, after penetrating the core, give an exposure at the
crew’s position of 7 X 10° Roentgens per day in the case of the assumptions
stated. A secondary heat-exchanger on the off-crew side of the shield proper,
assuming a factor of ten for shadow shielding and another factor of ten for
self-shielding, would still require a thickness of lead of 22 cmor an additional

weight of 17 metric tons for an exchanger a cubic meter on a side.

Bremsstrahlung from Lithium Beta Rays (N. M. Smith, D. K. Holmes). The
gamma-activity produced by the 12 Mev [-ray from the disintegration of Li® was
estimated. The method employed assumes that the Li is intimately dispersed in
an element of atomic number of 26-30 (the piping and radiators). The Bethe-
Heitler theory for the emission of 7y-rays was assumed correct. This theory es-
sentially assumes that a/8-particle makes not more than one Bremsstrahlung col-
lision in slowing down. Furthermore, the extreme relativistic formula was as-
sumed for the Bremsstrahlung cross section. The first assumption should result
in an under estimation of the y-activity; the second an over estimation. The

resulting estimate may therefore be in error by more than an order of magnitude.

There resulted a dosage of 30 R/day at the crew position as above, making
an allowance of a factor of ten for self-shielding and shadow shielding. While
this number should not be taken too seriously, it does indicate that Li’, if
used for a coolant, may have to be confined to a heat-exchanger and the latter

covered by light shielding.

Further work on this problem is contemplated either making use of a more
exact theory by M. E. Rose (ORNL 697) and the more exact Bremsstrahlung cross
sections, or by means of a Monte Carlo calculation. The latter is expected to

give approximate results with less calculation time.

RADIOACTIVITY IN THE EXTERNAL COOLANT OF THE CIRCULATING FUEL REACTOR
N. M. Smith

In the case of the circulating fuel reactor another effect might make the

external coolant radioactive even though it came from an intermediate heat ex-

18
changer in the shield and never passed through the core. The delayed neutrons
from the fission products in the inner circuit of the heat exchanger might re-
act on the dissolved fuel in such a way as to build up a high neutron flux
within the heat-exchanger. Thus even though the exchanger was far enough away
from the chain reacting core so that the ambient neutron flux in the shield
was not great enough to make the coolant in the secondary circuit (or its im-
purities) active, still the exchanger itself might be a strong enough neutron
source to activate the external coolant. Some examples of this effect are

calculated below.:

The primary fluid in the case of a circulating fuel reactor is assumed to
be a bismuth-lead-uranium alloy; the secondary (external) coolant being normal
lithium. The total power is taken to be 300 MW.* The heat-exchanger is assumed
to be constructed of Fe, with one-third of its volume each of primary and

secondary coolants.:

The theory of the neutron multiplication in the heat-exchanger resulting
from the presence of fuel and the emission of delayed neutrons has been
published as a separate report.(!’ It is shown that the slowing down density,

gq(r,7) in the bare intermediate reactor is given by:

 

q(r,7™) = Z F(n)P(B_,T) Z (r), (1)
where | J 0
F(n) = 7 = - (2)

| e | " a(r BB, T ) dr
- a(T LTy dT
1-kth eff(Bn)G

r position vector.
7+ = Fermi age.
P(B,,7) = the three-dimensional Fourier-transform of the point-

slowing down function from age 0 to 7 (including the resonance-
escape probability).

an, the buckling, 1s an eigenvalue of the Helmholtz Equation:

v?2 Z (r) * an Z (r) = 0 obtained by the boundary condition of
Z (r) finite everywhere and Z (r,) = 0, where r, is the vector
position of the boundary.

(1) Smith, N.-M.,The Bare Intermediate Reactor: the Approach to Critical, ORNL 665 (Apr. 18, 1950).

Megawatts.

19
et 2 ° .
b, = e "0Bn”, where Ty 1s the age at which the delayed neutrons
are born.

Q, is the weight of the nt" mode of the source distribution, i.e.,
Q(r) = Zfi Q, Zfi(f), and since the ZnCL) are orthonormal,

Q, = |  Z (1)) dr..

 

reactor
k  P(B 2,7 )
2 - th 9 - : . . . .
k.4 eff(Bn ) ———-—7—{t-1fi?5:) thermal effective multiplication
p\th 1 L n .
constant corresponding to the n‘? mode.
L = the diffusion length.
— O—‘fgh — . ‘e . . .
kv = Ten ‘ P,y © thermal infinite multiplication constants.
ath '
M, ~ fission neutrons produced per thermal neutron absorbed.
Tt h = thermal fission cross section per cmd®.
o +p — thermal absorption cross section per em®
_. 3 m(T) op(T)
a(r) = : .
AT
n(7) dr = fission neutrons produced per neutron absorbed of age

between 7' and TH#dT.

A(T) = transport mean-free-path as a function of age.:

Tf(TQ = fission cross section per cm® as a function of age.

P,, ~ thermal resonance-escape-probability.

The denominator of (2) 1s the negative of the excess multiplication factor
of the entire reactor. A very approximate evaluation of the excess multiplica-
tion factor of the fundamental mode for the heat-exchanger indicated to a first
approximation that the neutron multiplication in the heat-exchanger just
‘compensates for the leakage. The strong 1/v absorption of the lithium has the
property of cutting off the spectrum of neutrons before thermal energy is

reached, making the heat-exchanger a sub-critical intermediate reactor.-

The ratio, R, of the neutron absorption in one component of the materials
making up the heat-exchanger and 1its contents to the total absorption, was
computed by approximate means, using the neutron spectrum of equation (1).
There results that

- (03

= G g 3 3
T o) glyl, (3)

 

20
where Udi(O) is the fast macroscopic absorption cross section of the ith

component, and g(y) is a very slowly varying function of y, where

y = Tai (0) z Tai (Tfh)
(0)

Tai (Tfh) 2 Tq

 

i

g(y) is very close to unity. Thus a second simplifying result is obtained
that the various components of the assumed heat-exchanger absorb neutrons in

proportion to their fast absorption cross sections.,

Table 2 has been constructed assuming a power level of 300 MW, a chain-
reacting core volume of .6 X 10° cm®, volume of circulating fuel in heat-ex-
changer = .3 X 10° ¢m®, volume of lithium in heat exchanger = .3 X 10° ¢cm?®,
volume of lithium outside shield and exposed to crew’s position 100 feet away
is taken to be 14 X 10° cm®.

The concentration of various contaminants producing one Roentgen per day
at the crew’s position is tabulated. Note that normal contamination amounts
of all substances listed is permitted. At the same time, pure substances

(such as sodium) would produce radiation in excess at one Roentgen per day.:

One need not be much concerned with the (8,Y) emission from Li® in the

case of the circulating reactor.:

EVOLUTION OF FISSION GASES FROM A FUEL ELEMENT
D. K. Holmes, N. M. Smith

In order to examine the possibility that significant amounts of fission
products may diffuse out of bare fuel elements at 1000°C, the general ex-
pression for the current of fission product ions from the surface of a slab
inside whichP fission product iofis are produced per cc per sec has been ob-

tained.(2?) The accurate result may be approximated for small Dt by

. _ 2P
i =

where t is the elapsed time from the start of fissioning, and D is the diffusion

Dt , (1)

coefficient (cm?/sec) which must be obtained from experimental data. A series
of reports of the Ames Chemical Group(3-1%) give diffusion coefficients for

many fission products from uranium slugs, all ranging down from 1019 cm?/sec.

(2-21) A complete list of references is given on pages 29-30,

21
Gamma Emitting Contaminants—-Circulating Fuel System

TABLE 2

(Heat exchanger separating circulating fuel from external coolant.
Concentration leading to one R/day at crews quarters of y-emitter
of some common or possible contaminants.)

 

THERMAL CROSS -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FISSION ENERGY ATOMIC. TOTAL MASS (IF LESS
ABUNDANCE v*s EMITTED B° s EMITTED CROSS--SECTION| SECTION CONCENTRATION DENSITY THAN PURE SUB-
ITEM| ATOM (percent) REACTION (Mev) (Mev) HALF LIFE (millibarns) (barns) (Noz/cmg) (grams/cm3) STANCE) (grams) REMARKS
14 14 14
1 N 99.62 N" (n, p)C No ¥ 5700 Yy 40 Fast reactions in nitro-
gen are negligible
9 N4 99. 62 N'*(n, a)B! Stable Stable
3 NaZ3 100 Na23(n, y)Na2? 1.4, 2.76 1.39 14.8 h 0.29 0.6 2.4 x 102° . 0092 1.29 x 10°
4 Na23 100 Na?¥(n, p)Na?® No 4.1 40.7 s 0.7+ Neglect
5 NaZ3 100 Na22(n, a)F?° 2.2 5 12 s 0.4 > 3.3 x 10°° > .013 > 1.8 x 10°
6 Nal3 100 Na?®(n, 2n)Na?? 1.3 B" 0.6 2.6 vy - May be neglected
7 Mg2* 78.6 Me??(n, p)Na®? 1.4, 2.76 1.39 14.8 h 1.0 - > 5.1 x 10'° > . 0022 > 3.0 x 10%
8 Mg 5 10.1 Mg25(n, p)Na2® 1.7 3.4 61 s 2.5 . > 1.2 x 102! > . 046 > 6.4 x 10°
9 Mgzs 11.4 Mng(n, T)Mng 0.84 (100%) 1.8 (80%) 4.2 x 1021 . 170 2.4 x 106
1.05 ( 20%) 9.6 m 0.60 . 048
10 a1%7 100 A127(n, y)a12® 1.8 2.75 2.3 m 0.4 0.18 4 x 10%° .012 1.7 x 10°
11 a1?’ 100 A1*7(n, pymMg?’ 0.84 (100%) 1.8 (80%) 9.6 m 2.8 > 1 x 102° >, 003 > 4 x 10%
1.05 ( 20%)
12 a127? 100 A127(n, a)NaZ* 1.4, 2.76 1.39 14.8 h 0.6 - > 1.2 x 10%° > 0035 > 4.9 x 10*
13 5i28 92.27 Si28(n, p)a1?t 1.8 2.75 2.3 m 3.0 . > 5.8 x 1014 >92.7 x 107° > 3.8 x 10°
14 si2? 4.68 3i2°(n, p)a1?® ? 2.5 6.7 m 2.7 . May neglect
15 5i3° 3.05 5i3%n, ysid! No 7 1.8 2.83 h 1.1 0.12 May neglect
16 c13% 75.4 c1?%(a, a)p3? No 1.71 14.3 4 6.0 -
17 c1?’ 24.6 c1®*(a, yea1®® 1.60 (36%) 1.19 (36%) 37.5 m 0.80 0.60 .
2.15 (47%) 2.79 (11%) 9.3 x 102 -.056 7.8 x 10°
4.94 (54%)
18 k*! 6.7 K*(a, nK*? 1.51 (25%) 2.07 (25%) 12.44 h 2.9 1.0 1 —
3.50 (75%) 3.9 x 10° .27 3.8 x 10°
19 Ca No bad gammas combined with high cross sections or high abundance -
20 Mn>> 100 Mo3% (n, 7)Mn®® 0.84 (100%) 0.75 (100%) 2.59 h 3.5 12.8 N .
1.81 { 25%) 1.05 ( 25%) 5.9 x 10 . 0054 7.6 x 10
2.13 ( 15%) 2.88 ( 60%)
21 Fe>® 91.7 Fe*®(n, p)Mn°° 0.84 (100%) 0.75 (100%) 2.59 h 4.7 - 19 : > 4 -
1.81 ( 25%) 1.05 ( 25% > 4.8 x 10 > 0045 6.3 x10
2.13 ( 15%) 2.88 ( 60%)

 

 

 

 

 

 

 

 

 

 

 

 

 

* The application of this theory to fast threshold reaction is

expected to give an

upper Iimit to the radiocactivity.

22
This value for D makes Eq. (1) valid for times as long as 10° seconds, and a
‘simple estimate from Eq. (1) using P = 6.10'%/sec (corresponding to about
100,000 watts per cc) shows that even after 10° seconds only 1.5 cc of fission
products considered as gases at STP will have crossed a square centimeter of
surface.. Thus, the fission products will be effectively trapped inside the
fuel even at 1000°C unless (a) the fuel is in the form of a powder or a porous
compact, or (b) if the uranium itself suffers sufficient radiation damage that
it essentially breaks down into a powder. The Ames reports(3-1%) which
concern UO, are inagreement with (a), while recent GE experiments (unpublished

results) indicate that for one percent burn-up (b) is indeed the case.:

For the ANP, then, it must be concluded that essentially 100 percent of

the fission products will escape from the fuel itself during operation.:

The possibility remains that the fuel may be "canned." By enclosing the
fuel element in a container of, say, aluminum, the fission fragments which
come from the uranium on first flight are readily stopped in container walls
of a few mils thickness.(!®) Further, experimentally measured values for the
diffusion coefficients of fission products in aluminum and stainless steel(15:16)
are as low as those in uranium.. It must be concluded that even at 1000°C,
diffusion through undamaged container walls will not be appreciable,(17) so
that the fission products which do escape from the fuel itself may be canned
in, unless the effect of radiation damage, which may be toward increasing the

effective diffusion coefficients, (19:20)

becomes important at the proposed
high flux.- It is this latter possibility which stands in greatest need of
experimental work, since available theory is entirely insufficient for even an

estimate of effects of high fluxes on diffusion coefficients.:

ADVANCED SEMINAR IN REACTOR PHYSICS
N. M. Smith

An advanced seminar in reactor physics has been conducted for any interested
members from the Oak Ridge National Laboratory or NEPA; the meetings are held
once each week. The theory of the bare reactor has been covered in detail,

including:

a. Slowing down and diffusion of neutrons in the finite medium;

b. The approach to critical of the thermal reactor: the critical
conditions;

23
c.” Elementary kinetics of the bare reactor,
1. One velocity group, no delayed neutrons,
2. One velocity group, one delayed group of neutrons,

3. Bare thermal reactor with general slowing down kernel
and arbitrary number of delayed neutron groups and neo ex-
ternal source;

d. Same as c (3) but with an arbitrary internal source present;
e.. Kinetics of the bare reactor withuniformly increasing reactivity;*

f. The approach to critical of the bare intermediate reactor: the
critical conditions (ORNL 665);

g.. The theory of the oscillating absorber in a chain-reactor;

h. Perturbation calculations of kinetic response of a reactor as
influenced by various physical characteristics;

i.  Kinetics of the Homogeneous Reactor; **

j. First-order perturbation theory.

Copies of these seminar lectures are avallable on request.:

COOLANT DYNAMICS
J. F. Haines

Studies have been made of the influence of the physical properties of
potential coolants on the mechanical design of the nuclear aircraft power plant.
Li, Na, Pb and Bi have been considered as the primary choices, but results can
be interpolated for other materials by comparison of their physical properties

with those of Li, Na, Pb and Bi.

Table 3 gives the properties of the four coolants having the major effect
on the engineering design of heat exchangers, pumps and other coolant system
elements. It will be noted that the boiling point is comfortably high for all
but Na.  Pressurization at . approximately eight atmospheres might permit opera-
tion with Na at 1000°C; the vapor pressure of Na at 1000°C being about 2.5

atmospheres, and reaching only eight atmospheres at about 1200°C.:

The product of specific heat and density is the heat capacity per unit
volume.  This value isof primary importance since it determines the effective-

ness of the coolant in removing heat on a velocity basis.  Table 3 indicates

¢ Jecture by W, L. :Carlson

#* Lecture by J., M. Stein 24
TABLE 3

Properties of Some Potential Reactor Coolants

 

 

 

o SURFACE HEAT
MELT ING BOILING HEAT CAPACITY VALUES RELATIVE TO Li AT 1500 F TRANSFER COEF# .
POINT PQINT AT 15 OgF FOR A GIVEN GEOMETRY AND POWER h* AT 1soo;F PRICE,
COOLANT (°F) (°F) (Btu/ft”.°F) TELOCTTY SRESSURE o To (Btu/hr-ft“.°F) PER FT
L1 366.8 2928 29.8 1.00 1.00 1.00 30,900 $ 400
Na 207.5 1616 14.8 2.01 8.52 17.1 32,000 $ 12
Pb 620.6 2935 22.6 1.32 42.7 56.3 7,300 $ B84
Bi 519.8 2678 23.3 1.28 34.9 44.6 7,500 $1220

 

 

 

 

 

 

 

 

 

* Values calculated for flow through .1 in. tubes eat.a velocity of 10 ft/sec, -

25
the relative velocities, dynamic pressures and pump horsepowers for the four

coolants considered for a given geometry and power.-

From a practical standpoint, the limiting factor of the coolant system
and reactor structural design is probably reactor pressure drop, which will
account for the major stresses. In order to maintain equivalent stresses in
systems designed around the different coolants, it is apparent that coolant
velocities must be much lower for the heavy coolants, resulting in a large in-
crease in size of ducts, heat exchangers, etc. Designstudies indicate further
that in order to avoid excessive size of the reactor, with a corresponding
major increase in shield and structural weight, the choice of core construction

is much more limited for Pb and Bi than for Li and Na, The indicated superi-
(22,23)

ority of Li as a coolant is in agreement with previous surveys.

LIQUID METAL COOLED ARE DESIGNS
A. Fraas, R. Schroeder,
J. Wyld*

The ARE Design Group is attempting to achieve a 1000 kw reactor design
which will simulate as far as possible an eventual large liquid metal cooled
aircraft reactor. The first step in the process must of course be to develop
at least a rough picture of the large reactor itself. During the past quarter
a number of reactor arrangements have been sketched out which satisfy the
following set of general engineering specifications. These specifications are
believed to apply, in first approximation, to a reactor capable of powering an

aircraft at Mach 1.5:
1. Coolant - Li’,

2. Core pressure drop - 60 pounds per square inch,

3. Core size and shape - right circular cylinder 2-1/2 feet
height and 2-1/2 feet diameter,

4. Maximum temperature of fuel element in contact with coolant

- 1800°F,

5.- Power - 500,000 kw,

6. Coolant temperature rise - 400°F maximum,

7. Coolant outlet temperature - 1700°F minimum,

8. Structural working stress - 6,000 pounds per square inch,

* On leave from Reactor Motors, Inc.
(22) Lyon, R.:N.; Heat Transfer Experience on the Project, M-3974 (Aug. 5, 1949).

(23) Kitzes, ‘A, -S.; A Discussion of Liquid Metals as Pile Coolants, ORNL 360 (Aug, 10, 1949).:

26
o

9. Acceleration:load - 12 g vertical,
10. Maximum attainable moderator volume and minimum structural
poisoning,’

11. Thermal stresses assumed to be self annealing.

All of the designs studied during the quarter contemplate BeO as moderator
and UQ, as the fuel. Possibilities considered for fuel element designs in-

clude:

(a) 1/4 in. stainless steel coolant tubes coated on the outside with UO,
grains embedded in an iron matrix,

(b) Rectangular sandwich-type plates .025 in. thick, consisting of a
UO,-stainless steel ceramel layer coated on both sides with stainless steel,

(c) Small cylindrical stainless steel capsules containing possibly a
sintered powder of BeO plus UO,, or perhaps UO, plus Cr.

In each case molybdenum would be substituted for stainless steel 1if it
turns out to have a low enough intermediate neutron cross section.  Figure 2
illustrates some examples of the types of fuel-moderator arrangements being

studied.

All of these designs are faced at present with a large number of experi-
mental unknowns. In every case the feasibility of the reactor design will be
decided by materials questions for which the relevant data are yet inadequate.
During the next quarter experiments will be carried out in an attempt to settle

as many of the materials unknowns as possible.

NaOH--COOLED ARE DESIGN
H. K. Ferguson Co.

The Atomic Energy Division of H. K. Ferguson Co. is under contract with
the Oak Ridge National Laboratory to make an exploratory study of an aircraft
reactor operating with circulating liquid fuel. The contract covers a nine-
month period which started in March, 1950, The first two months of this
period were devoted mainly to a critical review of the available information
relating to the problem. For this purpose members of the Ferguson Atomic

Energy Division have spent considerable time at Oak Ridge, and information

27
 

 

 

]

=1

=y

 

BEQ MOPERATOR PLATES -

COOLANT FLOW PASSAGES
[ .O0850 " wipEe)

FLAT FUEL ELEVENTS—

- - .-|| ¢ - .-u._-

=T 1P REACTUR
o0 BD DRSS TN
Lk iz o0 s ar bl ad s

  

nwe. 9185

SUPPORT LANDS
FOR BEEL ELEMENTS

 

 

 

 

 

(.025" THIIK)

- |
"y B
——— - - —— -

 

Vv

 

_ CORRUGATED FUEL
L/ (.025" THICK!

 

ENDUS OF SHTS. BENT OVER & WELDED
REQ MODERATOR PLATES —
(CLAD WITH STAINLESS STEEL + B
BONDED TO PLATES) e 1 3
C"}C':__bh‘l- PASSAGES — —'_-j T 3
| <o
SPACER LANDS =X L as0%
ENLARCGED SECTIONS OF FUEL ELEMENTE b
FH e . { 1
W, = 1 )
ERAMEL SHEET 1.009° £ 7z 1
173 U0, 2/3 STAINLESS STEEL OO T A T Syt TR / T
8Y VOLUME (RONDED TO CERAMEL BY HOT-ROLLING
_'{?fi » J///r* A
..f'l- " 3
.008" STAINLESS STEEL SHEET -~ /7// //////////
(NEXT TO COOLANT) W
- e
009" CERAMEL SHEET—  /
(Y0, S:S:)
.w" OND | | A== T ~REQ
002" BONDING LAYER—
) MODERATOR J
(NICKEL OR IRON., BONCED TO BEO MODERATOR 0.2

8Y METAL SPRAYING OR VAPOR DE-
POSITION FROM CARBONYL)

ELEMENT ——

&

 

3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
 
 

PLATE
(0.162" THICK)

 

S A

 

28

 

 

 
visits were made to Battelle Memorial Institute and to Argonne National Labo-
ratory. In addition to informing themselves by reading reports, visiting the
experimental facilities, and holding discussions with various people at X-10,
Y-12, 'and NEPA, these personnel have participated in the ARE design conferences

held twice a week.:

As a first step in development of a liquid fuel reactor, Ferguson will
submit an ARE proposal. This proposal is directed toward the ultimate use of
an aircraft reactor in which NaOH serves both as the fuel solvent and the

moderator for a homogeneous circulating- fuel reactor.

Calculations were made which indicated that NaOH was satisfactory from
the standpoint of heat transfer. A preliminary survey has indicated that the
possible fuel-solvent combinations for a circulating fuel for high-temperature
use are limited in number, but this is at least partly due to lack of in-
formation. Inparticular, experimental data on the solubility of UO;, or other
uranium compounds, in NaOH are scanty and inconclusive. Such a system would
have important and obvious advantages. Consequently, Ferguson has requested

ORNL to make an experimental study of the pertinent properties of this system.:

The Ferguson ARE proposal comprises three phases. In Phase I, a heter-
ogeneous low power ARE reactor is used, comprised of fixed fuel elements (which
may or may not contain the fuel in liquid form) and circulating NaOH moderator.
The objective 1is to achieve as quickly as possible a reactor operating at high
temperatures, to gain operating experience and to test certain container
materials against molten NaOH under radiation. The heterogeneous low power
reactor is scaled up to a full-size reactor not by increasing the heat transfer
surface between fuel and moderator, but by eliminating it. This is done 1in
successive steps (Phases II and III) in one of which the surface is removed
making a homogeneous reactor, and in the other the fuel is circulated outside
at a rate sufficient to get losses in delayed neutrons. Depending on the

available uranium-bearing liquids, either of these steps may be taken first.-

Design work on Phase I is proceeding in New York.

GENERAL COOLANT CYCLE STUDIES

North American Aviation, Inc.

North American Aviation, Inc. 1is under contract with the AEC to perform

general analytical and materials studies pointed toward an eventual aircraft

29
reactor. During the past quarter the work consisted of a survey of various
possible power cycles for the aircraft reactor, in an effort to assess the re-
lative merits of the flowing liquid metal system as compared to a number of
vapor-cycle possibilities. A report on this work will be issued during the

next quarter.:

30
REFERENCES

Diffusion of Gases from U?35 and Mixtures

(2) Y-F10-5 D. K. Holmes, Nicholas M. Smith, Jr.,
Solution of the Diffusion Equation for Constant
Production in an Infinite Slab.

(3) CC-275 I. B. Johns, A. S. Newton, W, H. Sullivan, A. Voigt ,
Preliminary Report of the Ames Chemical Group on Dif-
fusion of Fission Products from U,0, and Sintered Met-
al (Sept. 23, 1942) .-

(4) CC-327 I.-B. Johns, A, S. Newton, W. H. Sullivan,
The Diffusion of Fission Products from Cast Metal at
600°C. and 1000°C. (Oct. 31, 1942).

(5) CC-342 F. H. Spedding, I. B. Johns,
Metallurgical Project Report, Section G (Nov. 15, 1942).

(6) CC-390 I. B. Johns, A. S. Newton, W. H. Sullivan,
The Diffusion of Fission Products from Polished and Un-
polished Metal at 600°C. (Dec. 14, 1942).

(7) CC-521 F. H. Spedding,
Metallurgical Project Report (March 15, 1943).

(8) CC-587 F. H. Spedding, I. B. Johns,
Ames Project Report (Apr. 19, 1943).

(9) CC-594 I. B. Johns, A. S. Newton,
The Diffusion of Fission Products from Cast Uranium at
1000°C. (April, 1943).

(10) CC-616 I. B. Johns, A. S. Newton, W. H. Sullivan, A. Voigt,
The Diffusion of Fission Products from Uranium at 1185°C
(May, 1943).

(11) CT-3765 J. E. Wilson,

Diffusion of Fission Products from Beryllium Fuel Rod
Material (Jan. 17, 1947).

(12) ANL-4185 W. M. Manning, O. C. Simpson,
Argonne Summary Report, Sec. 1.5 (Oct., 1948)."

(13) N-2049 0. Stern,
The Diffusion of Xenon in Tuballoy (July, 1945).

31
(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

Memo, HB-15 H. Brooks,
A New Problem for Pins {Mar., 1948).

Diffusion of Gases in Other Metals

BI-101 The Permeability of Metals to Argon and Helium
(Oct., 1946)."
MonT- 342 G. Ascoli,
The Diffusion of Fission Products in Aluminum
(July 14, 1947).
Diffusion and First Flight
BNL-C-2217 -~ H. W, Newson, |
Report on the Helium System Experiment at Oak Ridge.
BNL-20 J. Chernick, I. Kaplan,
The Escape of Fission Products from a Uranium Rod
(Nov. 3, 1948).
The Effect of Radiation Damage
TID-65. F. Seitz, v
( p. 142) - The Influence of Pile Radiations.,
AEC-500 J. C. Slater, D. S. Billington, |
Survey of the Effects of Radiation on Materials.,
KAPL-238 Technical Feasibility Report for the KAPL West

Milton Area Reactor (Feb. 14, 1950).

32
CRITICAL EXPERIMENTS

A. D. Callihan, K-25, J. F. Coneybear, NEPA

Organization of the newly formed joint ORNL-NEPA group to study aircraft
reactor design by essentially zero power critical assemblies is just being
effected. The construction of the Oak Ridge Area Critical Mass Laboratory,
Building 9213, is about 60% completed. The building is scheduled for occupancy
late in July. Provision has been made in the laboratory for two sets of
experiments to be carried on simultaneously. A part of the personnel in the
building will continue the joint K-25—Y-12 program of experiments designed to
yield data of value to the production groups. The other personnel will
participate in the ANP program. Early experiments in the latter will be
directed to procurement of design data for the Aircraft Reactor Experiment.
Requests have been made for the allocation of beryllium and uranium to the
program. However, in the event of delayed delivery of the beryllium, it 1is
expected that preliminary experiments will be done with graphite moderator to

test instruments and controls.

The efforts of the NEPA experimental physics section over the past few
years, directed towards criticality experiments, form a substantial background
for the inauguration of this ANP program. Assembly equipment for critical
experiments has been built and tested at NEPA and will be installed in the
new laboratory. The following paragraphs descriptive of this NEPA work are
reprinted from NEPA documents.

“The general arrangement of the assembly equipment is shown in Fig. 3.
The operation of the assembly table, which is now set up at NEPA, has been
tested with a 20,000 lb static load on the movable half which is driven by a
speed-controlled electric motor. [imit switches are located at intervals to
reduce the speed in increments as the movable part of the table approaches the
:contact position. No difficulty was experienced in the mechanical or electri-

cal operation.”

“Tests were started on prototype, safety, and control rods mounted on the
assembly table. Deflection tests were performed on the three-inch square
aluminum tube to be used in the honeycomb. This was done by loading a bank of
~tubes with lead bricks considered to be equal in weight to the maximum load

for the honeycomb. The observed deflection in the bottom row of tubes was

33
 

NEPA PHOTO C 00622

 

 

 

 

 

 

Pig. 3

 

 

   

ARTISTE CONCEPTION-~
CRITICAL EXPERIMENTS FACILITY
0.005 inch which is considered satisfactory. There was no apparent buckling
of the tube walls. An order has been placed for fabrication of a sample

section of the aluminum honeycomb.”

“The overall system is electrically controlled. Control and safety rods
have been worked out and tested in part. Control room equipment is being

assembled in a mock-up of the control panel.”

35
SHIELJING

BULK SHIZLDING EXPERIMENTS

C. E. Clifford J. D. Flynn
H. E. Hungerford H. W. Newson
E. P. Blizard T. V. Blosser
R. Tewis*
ileactor Technology {livision
Hesults for Experiments 5 and 6, 27% and 18% P& Ly volume in water,
have been re-evaluated and the data are shown in Figs. 4, 5, 6, and 7. These

are now believed to be accurate to within about 30% in intensity.

The data on the Fe-H,0-Pb shield for the Naval Reactor are also complete
and here included as Figs. 8 and 9.

Experiment 8, measurement of 33% Pb, 67% H,0 has been started. No data
are yet available, but extensive foil measurements have been made for the

first 10 inches of shield.

Much of the last period has been devoted to calibration of instruments
with emphasis on the gamma 1onization chambers. The latter have been standard-
ized in free air by means of a radium source, and this measurement agrees
with a calculated value within 20%. The absolute gamma fluxes as reported
in NEPA-1374-IPR-52, Fig. 39, should be reduced by a factor of 1.9 i 20%.
Uncertainties arise from the effect of the orientation of the cylindrical
chamber with respect to the source. Further work will be required for a final
calibration, but in the meantime data can be taken which will be internally

consistent.

All neutron detectors are now fitted with new type waterproof cases
and individual preamplifiers. The latter, it is hoped, will considerably

improve the internal consistency of measurements.

The monitor, which records the incoming flux from the pile, has been
replaced with a much larger B!% ionization chamber which is powerful enough
to run a Brown recorder directly, thus eliminating the necessity of an inter-
mediate amplifier and its corresponding uncertainties and servicing requilre-

ments.,

The positioning apparatus for the long (25 in.) BF,; counters has been
completely revamped so that uncertainties in position are now kept to about
1/16 in.

36

* NEPA
NV

G. 8835
10 = ] —
_\‘K—\I PbISIob =
— \ 2 Pb Slabs —
N FIG. 4 ]
\ -3 Pb Slabs EXPERIMENT 5 —
0= \ NEUTRON DATA 27% Pb, 73% H,0 =
E_ CENTERLINE MEASUREMENTS —
- 1 :
| =14 Pb Slabs ]
10 “‘:"‘ =
— \Ks Pb Slabs =
: N\
0= <6 Pb Slabs =
B X__ 7 Pb Slabs —
10° — \ =
'03: \ =
— &9 Pb Slabs —
E N\ :
To) IiPb Slobs"\%\ =
— \\\ =
- \\ =
IO :\\\\ —
- \\\ =
- t\\

10 = AANN =
— \\ ]

L ..
0610 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 18

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Z, SOURGE, Cm.
37

 
R PER HOUR

l!G. 8838

I |

 

R

| | IT]

|
|

Fig. 5 EXPERIMENT 5
100 GAMMA CENTER LINE MEASUREMENTS

27% Pb, 73% HL SHIELD MOCK-UP

 

 

I TTTH
L L

|
|

 

 

 

 

 

10— _
0= \\“ §‘5§ DETECTORS: ]
— a8 4 3 * o GAMMA PROPORTIONAL COUNTER —]
- \\\\\\\ Lo
B N \ 0 ¢ 10° ION CHAMBER B
° 0 \i\\\\\i\ s
E \‘ \ A\\QQ}\ ;
B 4\\ \ Q\ T [T 100%H0—
102 \x \ \\. \\\:; TS “\lpb

 

///

{ /
/]
o

L

TR

 

od] _/wl -/1
/)
L

)
/

J/
/

T
| L1

 

POSITION OF LEAD SLABS j

 

—1Pb 2Pb 3Pb 4Pb 5pb 5Pb 7Pb 9Pb 11Pb

N 20 0 " 3 K 7
L 00 19 | | | | | |
‘ 0 10 20 30 40 50 60 70 80 90 100 1o 120 130 140 150 160 170
Z — CENTIMETERS FROM SOURCE

38

 

 

 

 

 

 

 

 

 

 

EF——I
o
-SEenEh
DWG. 88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

) 36

o) =
?\\/l Pb Slab ;
B s =
| \ EXPERIMENT 6 —~

e \\ | NEUTRON DATA 18% Pb, 72% H,0 -
= X‘\z Pb Slabs CENTERLINE MEASUREMENTS =
[ WATER B

6 GURVE-

10 — \ 3 Pb Slabs ]

10 — =
— ‘ Pb Slabs |

4
10 = —
3 A\

10 — =
— \.,5 Pb Slabs —

o \\
§ 6 Pb Slobs/s\\ =
| K |
! 7 Pb k\

10 = Slabs—" Q\\ =
— 8 Pb SI b&\\ N

S -+ \\\’§ 5
[ \\\ —

Ial — \\\ —

10° \\
0 10 20 30 40 50 60 70 80 90 100 1O 120 130 140 150 160 170 180

Z Cm (Source)
39
RATE 7/HOUR

-2

10

=3

10

-4

0

-5

10

I!G.BBBQ

 

r 1

EXPERIMENT 6
FIG. 7

GAMMA CENTER LINE MEASUREMENTS -

SHIELD MOCK-UP

18% LEAD, 82% WATER

L

|

 

CTET

l

 

T

I

 

 

 

N

 

 

 

 

DETECTORS

O m GAMMA PROPORTIONAL COUNTER
O ® 10'2 |ON CHAMBER

A & 10'0 ION CHAMBER

 

[ 1L

|

 

LT

 

T

NS

a

N

U \\}\

 

 

 

 

 

L

|

 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

— ~ *\ _ 100%H20)
- 5\4 hl\g\ | 2
- ‘t\<swlpb -
= T~ 2P =
: J&H\ B
i | o
E J L.

 

 

 

 

 

 

 

 

| < _
- ' ~———6Pb __|
— -\\‘\“'L\\\ N\\\\~Jg,/,/f7pb ]
B POSITION OF LEAD SLABS — .
\ 2 3 4 5 & 7 2 s |
Eg § § § § \I 8Pb ]
0 20 30 40 50 60 80 90 Mo 120 140 160 160 170 180

 

40

Z —(CENTIMETERS FROM SOURCE)

 
RATE 7 HOUR

102

g4

=5

10

108

 DWG. 8840

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

| EXPERIMENT 7 ]
FIG. 8
GAMMA CENTER LINE MEASUREMENTS
Fe, HoO & Pb SHIELD MOCK-UP
[ Ol -
- N ¢ .
X \‘\\T\ )
i \‘\\ 0\ A\\
— Y =
E— . T\T\A\o\ \ ‘\ _
B A N
y _
j}\ ,\A\ | \“\
= N O T -
= ;\A\ \’\\//loo% H20 5
- ~ere AN 4 -
- \ ’ Al\ P\A\l —
° Qa\ N A~ ~—2Fe
= \\\é\\ Z
[ \l! j~—4 Fe _]
= 14 Fe 2 TYBOR>\ —
— DETECTORS ;
— O ® GAMMA PROPORTIONAL COUNTERS \f/'O Fe —
B o @ 10'? |ION CHAMBER _]
B A A 100N GHAMBER \é———lz Fe B
0 ¢ 108 ION CHAMBER |4 Fe
¢ & G.M. COUNTER
— 1 \w\QB\ 14 Fe + | TYBOR ~——
- l —
— ‘\’l ¢ l4Fe,IPb ]
- \‘\‘\4\ ' ' -
[ 1'\¢ I ]
| do_ ~e 14Fe,2Pb ]
- t—t—— 14 Fe, 3Pb ——
= —— TG =
— T\‘\‘I ~(14! Fe,4Pb
— POSITION OF IRON « ' | 14Fe,5Pb
% POSITION NOT MEASURED POSITION OF LEAD ’\’\’ |‘
4 »* »
SRS A e e rvaon ' o= =tere 50 —
§ §§§§ % § 9 N " §|3 ! §§ § (Plusr:3 l:r?ta'r;bgficgm ofltonk 10)l
NN NN prevent 'eakag
0 20 30 40 50 60 70 80 90 100 10O 120 30 140 50 160 170 180

 

 

 

 

 

 

 

Z - CENTIMETERS FROM SOURGCE

41
1

 

P

ad

| L]

\ FIG.9 — EXPERIMENT 7

 

LTI

NEUTRON GENTERLINE MEASUREMENTS
IRON AND WATER SHIELD MOCK-UP

L L

 

LT

 
 
 
 

L2 Fe Slabs

| L

\ -4 Fe Slabs

 

LT

|

 

 

   

 

 

 

 

 

| 6 Fe $labs DETECTORS
FOIL MEASUREMENTS BEHIND 14 Fe SLABS
COUNTER MEASUREMENTS BEHIND 14 Fe SLABS
FOIL AND COUNTER MEASUREMENTS IN PURE H20

| LN

B Fe Slgbs

 

® o b »

 

| 1T

 

FOIL MEASUREMENTS BETWEEN Fe SLABS

—_Aorfe Sta
§ Dz Fe|Slabs
b
X 14 Fe S
\

 

L

 

FITHT

NV

  

L L

 

RRRILL

L Ll

 

 

P

\

| [ L

 

T

Y /4

2 Fie Slabs

L L

4 Fe|Slabs

/

 

LTI

14 Fe |Slabs —

7
Y, A
% i

S|ob§
12 Fe|Slabs—{ 1

R

10 Re Slabg

 

LT

8 Fe Slabs/

 

T

 

 

 

 

 

POSITION OF IRON SLABS
2% POSITION NOT MEASURED

 

L Illllll/ |

| 2 8 » 10 %12 %14

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o

34567
e s
10 20 30 40 50 60 70 80 90 100 no 120 130 140 150 160 70 180
Z (CMS. FI:%M SOURCE)
A new air-conditioned instrument room is being installed, and it is hoped
that this will considerably improve the reliability of the instruments. Here-
tofore almost all delays in the program have been directly attributed to

instrument failure.

As a service to the MTR, gamma measurements around the MIR Mock-up have
been made by the group during the shutdown of operations incident to installa-

tion of the air-conditioned room.

The new fast neutron spectrometer, reported on elsewhere in this section,
has been installed in the tank and a few spectra close to the source will be
measured, The instrument is not sufficiently sensitive to measure at very

great attenuations in the lid tank.

In an effort to assist Brookhaven National Laboratory in setting up
their shielding program, the radioactivity induced in their tap water after
one hour’s exposure in the lid tank thermal flux of 10® was measured. The

preliminary results were:

Distilled H,0: .010 microcuries/dc

Brookhaven water: .016 microcuries/cc

INTERPRETATION OF LID TANK DATA
E. P. Blizard

Reactor Technology Division

Thus far only thermal neutron fluxes have been measured in the 1lid tank.
Since intensity limitations make difficult the measurement of fast neutrons
directly through the required attenuation, it 1s necessary to develop a re-
liable method of estimating fast neutron leakage from thermal flux. Since
there is no resonance capture in water, this medium is ideal for correlation

of fast and thermal neutrons, by what is referred to as neutron accountability.

The power of the source plate has been measured by means of thermo-
couples to be about six watts. From this value the fission rate in the source
plate is known and with a guess about the reflection coefficient, one can
estimate the total number of neutrons entering the tank. This number agrees
well with the integral over the tank of thermal flux times thermal absorption

cross section.

43
A simple derivation yields a relation between location of removal of a
fast neutron from the incident beam and its appearance as a thermal neutron.

This displacement, 8, is found at large distances from the source, to be:

where 7 is the Fermi Age, and A is the relaxation length of the fast current.
Since & is a slowly varying function, A’'s for slow and fast neutrons will be

about the same.

The relaxation length is crudely found to be just the reciprocal of the
total fast cross section, and from this one can find, again crudely, a domi-
nant energy of neutron for any relaxation length in water. The pure water
data so analyzed show a reasonable approximation to the fission spectrum, as

would be expected.

From the energy associated with the relaxation length one can find
calculated values of the Fermi Age, 7, and from this obtain the displacement
§. T is chosen for the energy before collision, which gives an overestimate,
but on the other hand the scattered neutrons are distributed in the forward

direction, which probably compensates adequately.

Neutron accountability demands that for plane geometry,

 

L T ) 25, 4, G
where
2z =z + 9%,
I (z) = fast neutron current at z,
2, = thermal macroscopic absorption cross section, H,0,
¢ o (z') = thermal neutron flux at z'.

It follows directly that

I(;) FNZ, & (37).

Thus, for a given measured ¢th and A at z' one can estimate the fast neutron
current passing the plane at z. Since back-scattering in water is not great,
this then gives a measure of the fast neutron current which would be observed

with a shield ending at z.

44
This method has been applied to some of the shield mock-ups that were
tested in the 1id tank and a weight estimate for an ideal unit shield (no

ducts, no shield cooling, no structural members) has been made.

For purposes of calculation the following values were used:

 

 

 

 

 

 

 

REACTOR LID TANK SOURCE PLATE
Power 4 x 108 watts 6 watts
Leakage of fast neutrons 20% 25%
Size 2 ft radius (sphere)| 14 in. radius disc = 4000 cm?
Reactor-crew separation = 100 ft
Tolerance: It is assumed that the military dose i1s eighty times that allowed

at the Laboratory.

LABORATORY (8 hr day) MILITARY (25 hr mission)
Gamma dose rate 1/80 R/hr 1 R/hr
Fast neutron dose rate 50 n/cm?/sec 400 n/ecm?/sec
Note that the dose can be taken either in gammas or neutrons, and if

both types of radiation are incurred that the sum of fractional doses should

not exceed unity,

The following symbols will be used:

r = inner radius of shield
ro - outer radius of shield
H(r0 - r) = attenuation of spherical shell shield
F(z) = attenuation of plane shield, thickness z
h = factor to apply to lid tank centerline data
to convert to plane geometry
a = lid tank source plate radius .= 14 1in.
S = source strength, neutrons/cm?/sec

A5
We use the following geometrical factors:

H (r, - r)‘:_z_ F (ry, -r)

o

 

2 A

a

_ 2A? z 3
h = |l —+ 1| + % (See ORNL 629, p. 10)

Ratio of surface strengths of lid tank source to reactor:

[ 6 x 0.25
Spp — 4000
— = = 2.1 x 1077
S, {4 x 10® x 0.20 ]
4 7 (60)2

 

In order to minimize the heavy gamma shielding required, it is decided

to accept 75% dosage in gammas and 25% in fast neutrons. Slow neutrons are

assumed negligible.

A distance z' in the lid tank is then chosen so that the thermal flux 1is:

X 1000 n/cm?/sec

 

1 r, {100 x 30} . 1]
¢th (z') = —h—X—X‘ —_ Xx2,1 x 107" x S

r Eq

a

where the factors in order are:

1) correction of 1id tank centerline measurements to plane geometry
2) correction, plane geometry to spherical geometry

3) allowance for reactor-crew separation

4) allowance for lower specific source strength

5) conversion factor, thermal flux to fast current

6) 25% of fast neutron dose rate,

46
 

Choose & conservatively at 10 cms, and from Fig. 4 for 27% Pb and 73%
HZO, it is seen that z' = 128 cms satisfies the above on the curve labelled

11 slabs. The shield thickness determined by neutrons is then 118 cms.

The next assumption is that the emergent gamma rays all result from
neutron scattering or capture within the shield. It is then required that the

gammé flux at z = (128 - 10) cms must be measured in the lid tank to be:

1 r (100 x 30)2

‘r‘ ey — X ———————— x 2.1 x 1077 x .75 R/hr
LT h 2
s

. .018
I Roentgens/hr=4.5% 10°° R/hr
0

These requirements are satisfied at z = 118 cms by the estimated value
for about 10% slabs (Fig. 5). 95 cms of matched shield, plus 23 cms of H,0

are adopted.

Note that the second lead slab does little or no good for gammas. It 1is

neglected in the weight estimate.

Weight estimate:

4 4
W 3?77 (1783 - 603) +§—77 x .27 x 10.3 (1553 - 603) grams

= 64 long tons, or 70 short tons.

47
It 1s to be noted that in the measurements no advantage is taken of boron
to reduce capture gammas. This would probably introduce a saving of 5 to

10 tons.

NEW BULK SHIELD TESTING FACILITY

W. M. Breazeale
J. L. Meem*

Reactor Technology Division

Formal approval for construction of the new bulk shield testing facility
has been received, and the period since the last Quarterly Report has been
used in completing mechanical design work. Since the construction i1s to be
done by an outside contractor on a lump sum bid, i1t was necessary for the ORNL
Engineering Department to prepare complete plans and specifications for the
building and pool. The low bidder was Jokn A. Johnson and Company, and work
began on May 25. The contract calls for comuletion within 150 days. The
reactor, supporting bridge, and the bridge for the measuring equipment have
been designed and will be built at Oak Ridge or under subcontract. Schedules
are such that this equipment will be ready for installation when the building

and pool are completed.

Figure 10 gives the fundamental dimensions of the facility and indicates
the possible reactor locations. Details of the reactor design, which essenti-
ally duplicates the MTR design, are shown in Fig. 11. The fuel elements
(viewed end-on in section A-A of Fig. 11) consist of curved plates contained
in aluminum boxes and running the length of the core. The curved plates are

U-Al alloy clad with aluminum.

A three-dimensional sketch of the pool, reactor, and reactor bridge
is shown in Fig. 12. A similar bridge (not shown) will support the measuring

equipment.

It is still too early to predict accurately when the facility will go
into operation, but allowing a reasonable time for installation of equipment,
calibration of instruments, etc., and adding this to the contractor’s time
for construction of the pool and building, it appears that investigation of

the initial sample will begin during the first quarter of the next year.
& NACA

A8
 

-4 Ry o 274 _ o o
' ’ zv-1t
‘ - T T
17 i i
v i i
* —_ |
;
|
e !
. ! "*\L
T
v ) I
Ners!
‘ b For Appimoury DIMENSI0NE § SecTions SEe Dwe TD 1844
v » 1 !
v - 1
D T, 1 93
v TR ;’ :{') l} |
. LR ; * | i
B L - ]
Y Trun§ Rercron
CLosES T APFROACH ~Z'J bVl A
T
v v
Yol
) J
' _r "‘
, '
. .
. P .
’ I H “
. .
’ -
-
e - .
- C
v E xRxwwum.n Acumisivm BARRiER ’
o . L s For Disusing Orm Secrioss Ow Poon
© . 3 2 “
. 2§
' P
. i §
. 3 § :
' ?
’ 5 i 3 .
v z é F -
* E ‘E
f W g ")
’ *
M ¢ e i \: .
v ¢ g ;& ]
. s 24 ]
s &
? ! . . 7 RemcTow Max Beiocaruu b Tais
7 ? A 1 SECTonm OF Poel Fou, tARASUHEMEIN S
» o t' v [ Orey Wes T e Tien 1s Fiuceo -
. b —{'zsr AuRare i—L . _\‘/ Wit WkTER
¢ . ol i . -
T-e% I . I |
: | 4 o
‘ B i ‘ 8 h Reacvom . e ; bl Runcror Apavsrasie Avss Trog oy
- - T P « T S Nhvan 3, Twar e Evtes WL w - -
r7 b 3 ji i i l
1] ; ‘ i .
' o H“’""‘fl ST
. = Ll':: 2l C
c ¢ .. . v —- Rerqounone Hicn DEnsiry Conmire Buotrs
T /
» | -
. i - - | S
' - P s . T . - L L. - 4 T - . Lo I co - - - DO MOT SCALE THIS DRAWING.
. A ~ .
v o ’ oo a0 ot - - -t ‘ arpo. Toat: JOAK RIDGE NATIONAL LABORATORY
© » ’ . ° 2 » - « . L . - - , « | o _ —-7 — e
: . E . - te, ‘ ' £:4¢ *F TEGHNICAL DIV.—— P O. 80X P—— OAK RIDGE, TENN.
<. " ; T pa— Fig. 10
- ‘ A .. S - 3C10_RBRLDG
» . > L . - TR T T
‘ . » - - . [
— — P BULK_SHIELDING FACILITY
E . « YL - - -
» - o om . - s . -
e e . S S . . .
N\ . rl ! . . L |
\ ’ - - + 1 f A |\ Was 3= et |e-zz
e NS e — - ! jvev o REVISION TAPPC. | OATE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

etanear - weirasTOws M ¥
os

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L
Ll
1

 

 

i

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

31

 

30

 

 

ReMovmsle Rar.gctor

/—Fvu Asstveny

7‘ Skrery Raos
<

 

 

0000

 

 

 

) -
. ¥

g
a =

A
/

,,/

,

 

 

 

 

NN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Awove Tims Secvion

\-Cowrwm. £ Saewry Crammmrs

 

 

I
onrroL Roo

Sum -Fium:ur(BuO\R-lumoa

Section AA

RESTRICTED DATA _

desument contains re
mfl-&:  Samic Enercy Adof I8,

CAUTION

This docume!

inch

1ts transmission of the i horived

in any manner to an unuy ere criminy
y result in sV
bibited 238 Tovicable Federal laws.

-

-
I -

-

nt contains information
i Stutes.
National Defense of the ij:;d;";" . d

-

nis

 

 

 

TEGHMICAL DIV.—™—PR0. BOX P~ OAK RIDGE, TENN.

 

 

 

 

SHIELDING FACILITY

 

 

 

 

 

 

 

 

 

 

 

DN .
i [; N\ N
il . r ;
i ¥ ’ )
i . T
A o . ]
; ‘ T
<
2 b *
z S
9 o |
3 r :
»
£
3 . 3
I \0'-g" Res . "
o . -
é . . 4 REACTOR . View B-B
3 r ; - | —_—
¥ - O e
v - - -
o A =y 4l A * %% of
« ol "?1 - — —1_!_
H B :5 1-8
Ry |'—\O'
L L T T S e oo ’ Nova!
. -« ~ . - < N - “« |- v « <«
- . . 7 - |~ - - |- = - o = 1. Sucnion A-A Suows Tyecal Loapine. ACTUAL Loabing WiLL Dresno
o < o e = = e o< 5 v alaa )T < sle o “ On Tasr CondiTions
- : «~ - e = L e o " - -7 ol 2, Fom AbobiTionaL SecTions Sme Dwa TD (845
— ~‘ : * = ‘; = = DO NOT SCALE TWIS ORAWING.
e ~ oo s B oty
s e T e s LD e ] [An7o. ATE RIDGE NATIONAL LABORATORY
el S Sel ol < - Wi e s
‘- - : - s . s o[ = - «
R Tk : 3010 BLDG
L« s .- e “1. oe N, .
i‘_:r‘- ‘ :; . . ;’ » ’, " ’ . :: . ) /r.'_‘ F.g. ll BULK
T . : S T FUNDAMENTAL DIMENSIONS
. O A . . | ¢
s . s : . _ [oRKGaw o Joare
- < “ - B Gl &~15-5o
14’0
] T 0 8Y DATE
209 ng. GELAWM 3-0" | G l 6-22-50

 

 

 

 

 

 

 
 
With the relatively high intensity from the reactor, it is hoped that
measurements can be made of neutron and gamma ray spectra through shielding
samples in addition to the thermal neutron and total gamma ray ionization
measurements, such as are being made in the lid tank. A neutron camera similar
to that used at Los Alamos is under construction. Neutrons are collimated
through a two foot pipe, one inch in diameter, and strike a paraffin radiator,
ejecting protons. Experimentally, thelos Alamos group has found little change
in the neutron spectrum above 1 Mev upon collimation through such a pipe. The
protons are collected on a plate coated with a proton-sensitive photographic
emulsion, the plate being placed at a given angle with respect to the neutrons
incident on the radiator. The plates are examined under a microscope after
development, and from the number of proton tracks and the length of the tracks,

the intensity and energy of the neutrons may be calculated.

Several other instruments suitable for either neutron or gamma ray
spectroscopy are being considered for use in the new facility. It is hoped
that the feasibility of such devices will be determined by the time the

first shielding sample goes in the pool.

SHIELDING ANALYSIS

W. K. Ergen* F. Murray
K. Keyes S. Podgor*

Reactor Technology Division

Attempts are being made to summarize shielding theory, and to express
every step 1n as simple a form as possible. This 1s a long term project, but
some items are made available in the form of memos as soon as they are de-

rived,

In CF 50-4-3, identical with NEPA STEM-57, the neutron spectrum emerging
froma water shield was computed, assuming an incident fission neutron spectrum
of the form e'0°72E, and a total molecular water cross section which can be
fitted by o(E) = 11.6 E-'/? barns, in accordance with the H cross-section data

of B. E. Watt, [A-718, and 0 cross-section data of H. Feshback, ORNL 417.

The emerging neutron spectrum has a maximum, since neutrons below the

maximum are attenuated strongly by the large cross sections, and neutrons with

* NEPA
energies above the maximum are not very abundant in the incident spectrum.
The maximum moves to higher energies with increasing shield thickness because
the attenuation, which 1s less for higher energies, becomes more important

than the abundance in the incident spectrum,

Similarly, the neutron spectrum emerging from an iron water shield was
computed in CF 50-3-108 (NEPA STBRM-55). It was found that the emerging
spectrum for 2/3 iron by volume, as used in lid tank'Experiment 7, has a peak
at about 1 Mev because of the cut-off in inelastic scattering in iron at low

energies.,

Since the fabrication troubles with metallic wolfram sheets proved to be
greater than anticipated, the advantages of W for shielding were reexamined in
CF 50-3-103, NEPA STRM-53. The high density of W gives savings in shield weight
which amount to 10%, as compared to lead, in representative cases. This 1is
partially offset by the lower Z,Ewhich_makes W a less efficient gamma shield-
ing material than lead. Both the density and the Z effect are easily and
accurately computable and would not justify separate W experiments. The un-
known shielding properties of W are connected with the numerous and low lying
levels of W, which give hope of cascading of capturg and inelastic scattering
gammas, and hope of a low inelastic scattering threshold. Information on some
of these points is, however, appearing in the open literature. All-in-all,
experiments with metallic W appeared not worth the difficulty in fabrication,
and it was recommended that fabrication of metallic W sheets be abandoned for
the time being. (As reported elsewhere, W experiments are expected to be
carried out with a more easily obtainable material, such as W powder bonded

by tygon plastic,)

Among hydrides, UH, seems still worth investigation as to obtainable
density, reduction of pyrophoric behavior, and filling of interstices with

useful shielding material (CF 50-3-107, NEPA STRM-54).

Another interesting hydride is LiBH,, which probably can be cast and
hence obtained in theoretical density. Its main advantage would be the low
weight of compound per gram atom of hydrogen, 5.4 gm as compared to 9 g of
water, LiBH, of theoretical density also contains somewhat more hydrogen
atoms per cc than water, and Li and B are better neutron attenuators than

oxygen.,

Titanium hydride is not too goed as shielding material but holds the

hydrogen at high temperature. Some attempt should be made to obtain TiH, at

53
theoretical density, or to fill the interstices.

These comments on hydrides were set forth in somewhat more detail in

CF 50-5-24 (NEPA STRM-63).

Liaison was maintained with the Westinghouse-Argonne and the KAPL Naval
Reactor Shielding groups and one of the conferences was summarized in CF 50-3-6

(NEPA STRM-52) .

As a service to the MIR group, the expected relaxation length in Barytes

concrete was computed (CF 50-4-54) and found to be about 8 cm.

The analysis of the penetration of fast neutrons through thick shields
with the Boltzmann equation was further developed to include an approximate
calculation of the contribution of the continuous spectrum, for the special
reduced form of the Boltzmann equation in which the only scattering assumed
is elastic scattering from heavy elements, and possibly others considered

as pure scatterers.

NEUTRON ENERGY SPECTROMETER
B KR Gossick*
K. Henry*

Reactor Technology Division -

During the past quarter the fast neutron energy spectrometer was tested
at Bartol Research Foundation and at Massachusetts Institute of Technology.
The electrostatic generator at Bartol was used to obtain 4.4 Mev neutrons from
the d(d,n)He® reaction. Points at 1.9 Mev and 14 Mev using the Li’(p,n)Be’,
and H3(d,n)He* reactions, respectively, were obtained on the Rockefeller gene-
rator at MIT. The resolution checked with design calculations, giving on the
average a 23% half level width, which is considered sufficient for shielding
measurements. Results of the tests at Bartol and MIT are given in the joint

reports, ORNL 711 and NEPA-1407.

DUCT THEORY
D. W. Whitcombe®

Mathematics Panel

The effect of shield ducts upon flux distribution is very important when

a shield must be designed for minimum weight. Theoretical results have been

* NEPA

54
obtained for some simplified arrangements.

A diffusion solution has been obtained for a cylindrical duct with walls
perfectly absorbing thermal neutrons, the duct being filled with a pure
scatterer. Laplace’s equation is solved for a plane source of thermal neutrons.
The solution is found in terms of the extrapolatéd boundaries r, and z,.
, and z,. A condition on the
diffusion analysis 1s obtained which requires r, to be greater than 1.6 A

These are then related to the actual boundaries r

1
where A is the mean-free-path. When r < 1.6 A the first collision density

should be used. A detailed report of this problem is being prepared.

Another problem, diffusion for a cylindrical air-duct for infinite

geometry, has been solved by two methods. The first solution in an iteration
solution; the second uses the steepest descent approximation. Both methods
solve the same diffusion equation and boundary conditions. The results have

been published as ORNL 668.

Some transport equations have been set up and the matrix elements are
being computed so that the solution can be found on the NEPA computing machine.
Some work has been done on the three-dimensional cylindrical duct and 1t 1is
hoped to reduce this transport integral equation to linear equations so that

it may be solved on the NEPA machine.

SHIELDING SURVEY

Nuclear Development Associates, Inc.

Although the NDA contract for shield analysis is supervised by the AEC
Washington Office, The Oak Ridge National Laboratory maintains a strong interest
in those phases of the work which are applicable to ANP reactors. The program
involves study and application of analytical and numerical methods for pre-
dicting shield performance, study of results obtained from transmission and
bulk-shielding experiments, and work on problems arising in moving toward
actual shield designs, with liaison on engineering and materials problems.
In general the work is along lines indicated in ORNL 437, Summary Report on
Summer Shield Work, by Gale Young, with increased attention to practical

design problems encountered in mobile reactors.

593
‘SHIELDING MATERIALS

A. S. Kitzes
V. L. McKinney

Reactor Technology Division

Two additional attempts were made to roll Boral into large sheets at
Lukens Steel Company. One sample was heated to 1050°F and the other to 1150°F.
Both attempts were unsuccessful, probably due to rough handling since the
mill used 1s designed for steel ingots and the rolls could not be slowed to

the desired speed.

At X-10, twelve ingots 4 in. X 4 in. X 1 in. with B,C content varying
from 20% to 50% by volume were successfully rolled into 1/8 in. sheets. The
ingots were heated after each reduction. Initial temperature was 1100°F and
the rolls were lubricated with kerosene. Two ingots containing 50% B,C were
successfully rolled to 1/8 in. without heating between rolls by completing
the rolling within 3-1/2 minutes.

Two ingots are now being cast for rolling into sheets 24 in. X 84 in. X

1/8 in. Methods of joining Boral sheets are also being investigated.

56
HEAT TRANSFER

C. P. Coughlen H. C. Claiborne
A. R. Frithsen* R. N. Lyon

Reactor Technology Division

Experimental heat transfer equipment is now nearing completion. The
system, Fig. 13, is a figure-of-eight apparatus with liquid passing from a
pump through a preliminary heat exchanger, to the test exchanger, and from
there to a heater. The heated liquid is sent back through the test exchanger,
countercurrent to the cool liquid, then through the preliminary exchanger,
and finally through a cooler, before being returned to a sump attached to the
pump inlet. By-passes are built around both streams in the test exchanger, so

that flow rate on one side can be varied without changing flow on the other

side. Calrod tracing of all components permits preheating of the system to
prevent premature freezing of the lithium charge. Piping work is essentially
complete; Globar mounts have been constructed in the heater. The main heater

control and control board have been constructed and installed. Clean out runs

and pressure testing are scheduled for the near future.

Theoretical investigations of shape effects on local heat transfer
coefficients have progressed to the point of locating velocity distribution
data on rectangular and triangular tubes. Attempts will be made to interpret

these data in terms of local heat transfer coefficients.

A level indicator for flow rate determination using a catch tank has been
developed. A small amount of additional work is required, however, to make this

indicator operate a timer.

To date, pump development for experimental systems has been restricted to
the development of hydraulic bearings as a means of eliminating seals in pump-

ing systems.

The hydraulic bearing presently undergoing tests is a modified type of
journal bearing which derives its load carrying capacity‘from high pressure
fluid which is introduced through small orifices. After passing through the
orifices, the fluid pressure isdistributed longitudinally along the pump shaft
by means of eight slots; this produces a constant pressure around the shaft

periphery as long as no transverse loads are experienced. When transient

* USAF 57
SECRET

CoOLD BY

FASS

EXTHANGER

e g
TEMPERATURE

CONTROL
HEAT EXCHANGER

 

 

lf-

BY" PASS

 

| o

 

 

 

 

]

\

§

\

y M .'.‘.‘_\L"‘_l". IT

 

D AP
i

TANK

CO0

58

;'lfi'j -

9186

 
transverse loads are introduced, the shaft tends to move to one side of the
hydraulic bearing, and the flow from that side is reduced which, in turn, in-
creases the fluid pressure on this side but reduces the pressure on the opposite
side. Due to the pressure differential then opposing the transverse load, the

shaft tends to move back to its central position.

Figure 14 shows a complete pumping system, excluding pump impeller, which
has been set up for the purpose of testing hydraulic bearings. The pump
impeller is omitted in all tests in order to eliminate pump parameters from
influencing data obtained for evaluating the bearings. It should be noted
that an enclosed ("canned") rotor motor 1s used for the pump drive, which,
when combined with hydraulic bearings, not only gives a pumping system void of
seals and ball bearing, but also one that can be completely welded together

thus becoming leakproof.

During the next quarter, approximately 10 different variations of this
type of hydraulic bearing will be tested. 1In addition, one bearing of this
type has been made entirely out of transparent plastic so as to observe, with

the aid of dyes, the hydrodynamics involved in its operation.

Editing of the Liquid Metals Handbook, written jointly with representa-
tives from other AEC and Navy Laboratories and contractors, 1s now essentially
complete, thanks to invaluable assistance by the publications staff at the
Naval Fesearch Laboratory. The final proof will be submitted to the Govern-
ment Printing Cffice by June 17.

59
 

Y=-12_PHOTO 6-1446
NOT GLASSIFIED

 

 

 

INDUCTION MOTOR STATOR,
Af meui.&asmms

.
- '._‘_-_.-—
a » "

STATOR GOOLANT INLET.

           
 

. ROTOR

QAI.'H

ROTOR "CAN" FLANGE

Qf_up_usmsz N
) _ -\

4
» \, 4
=

v

  
  
    
    

J .",- r ‘_ TOR
*C GOOLANT QUTLET

..

.!,__

»

 

         
  

"THRUST BEARING

ON MOTOR ROQTOR

MP IMPELLER POSITION

 

 

|LIQUID METALS PUMPS

 

 

 

 

 

Fig. 14

 

 

 

60
EXPERIMENTAL ENGINEERING

H. W. Savage

The . objective of this group, newly formed, will be to provide data from
engineering experiments to determine the validity of design conclusions and to
test the components of the Aircraft Reactor Experiment. Ifi will also apply
the findings of other groups performing fundamental studies to full scale

engineering tests in components or mock-ups of the reactor.

Among the components anticipated to .be developed, tested and evaluated
are the fuel, moderator, container and reflector designs of the ARE assembly;
the primary and secondary cooling systems including the heat exchangers and
the required instrumentation; the internal and external power control systems;
the components of the shielding system; and equipment for handling components

remotely.

This activity will be initially centralized in Building 9201-3, Y-12.
The group will work collaboratively with the ANP groups engaged in fundamental
studies, and will be closely coordinated with related activities carried on by

NEPA.

61
METALLURGY AND MATERIALS

E. C. Miller

Metallurgy Division

The OENL Metallurgy Division is engaged in a materials program whose

objectives are:

(1) the selection and study of materials suitable for the

construction of a high temperature, liquid metal cooled, airborne reactor-—and

its prototype, the aircraft Reactor Experiment; and (2) the development of

methods for fabricating these materials into reactor components.

This program includes:

(1)

(2)

(3)

(4)

Statilc corrosion sorting tests in whichawide variety of possi-
ble structural materials are being exposed to several potential
coolants for different times and at different temperatures to
select the more promising combinations of materials for further
investigation.

Dynamic corrosion tests, using

(a) thermal convection loops to investigate the effects of
thermal coefficients of solubility on corrosion in a
circulating system subjected to a temperature differ-
ential,

(b) forced circulation corrosion loops intended as simu-
lated service tests, and

(c) spinner tests to investigate erosion in small scale
tests at constant temperature.

Investigation and control of observed static corrosion phe-
nomena and the factors affecting corrosion. Some of these
include surface finish, crystal structure, presence of pre-
cipitated phases, form and concentration of alloy components,
stress-corrosion, addition agents and gettering agents, mass
transfer effects.

Fabrication of reactor components and associated equipment.
This includes composite fuel elements, moderator matrix, and
heat transfer equipment.

STATIC SORTING TESTS

The tests now available for reporting consist of exposing the metals

listed in Tables 4 and 5 to the action of liquid lithium, and to bismuth,

62

in
 

 

 

 

 

 

& » - o ( ®
TABLE 4
Corrosion of Materials by Liquid Lithium in Static Tests - 4 hrs , at 10066°C
RATIO OF
2 SPECIMEN AREA
WEIGHT CHANGE /
GE (mg/cm”) TO COOLANT
» VOLUME,
MATERIAL SOURCE AND ANALYSIS RUNS RANGE AVERAGE cm? /em3 REMARKS
Titanium Remington Arms 3 + 3:5 to + 4.4 + 3.5 0.4 Discontinuous adherent film.
Ident ified by X-ray as TiN.
Zirconium Bureau of Mines 3 + 0:3 to + 0:5 + 0.4 0.6 Thin continuous adherent film of
ZrN-
Molybdenum Fansteel 3 + 0.1 to + 0:-4 + 0.3 0.6 MoC and H02C detected
Zirconium Bureau of Mines 2 - 0.1 to + 0.1 v 0.1 0.4 ZrN, same as Zr above
Niobium Fansteel 3 - 1.9 to + 0.1 - 0.1 0.2
Iron {(Afmco) Corvy Steel Co. 3 - -2 io - 0.1 < 0-1 866 (2} Uniform attack
Tungsten Fansteel 3 = 5:6 ¢o = 01 - 0.2 0-6 Uniform aittack - no film detecited
Tanialum Fansteel 3 - 1.5 to - 1-0 - 1.5 0.6 Uniform attack with thin carbide
fiim formed
1040 Steel SAE Steel Co. 3 - 2:4 to - 1.9 - 2-2 0.6 () Selective type corrosiomn, probable
carbide solution
Cobalt Kulite Tungsten Co. 2 - 5.2 to -~ 4.8 - 4.8 0.6 Penetration attack of sintered
metal
Chromium Univ. of Cinc innati 1 - 5.7 0.7
L-605 Haynes Stellite Co. 3 - 7.4 to - 6.0 - 6-8 1.3 Selective type corrosion
Vanad ium KAPL 1 - 7.7 0-4 Cont inuous non-uniform film of VN
and V_N or V_C
2 2
Beryllium 3 - 18 to - 16.4 - 17 0.6 Unidentiféed discontinuous sur face
film., Fairly uniform attack
Thorium Iowa State College 1 - 67.-7 to - 45.4 - 56.8 0.6
Nickel International Nickel 3 - 100.4 to - 67.9 - 86 0.6 Severe intergranular attack
Manganese Electro Manganese Corp. 3 Thin stcip tested - some solubility noted.
Silicon 3 Dissolved and deposited on capsule wall.

 

 

 

|

 

 

63
TABLE 5

Corrosion of Materials by Bismuth in Static Tests - 4 hrs. at 1000°C

 

2
WEIGHT CHANGE(mg/cm™ )

 

RATIO-—SPECIMEN AREA
TO COOLANT_VOLUME .

 

 

MATERIAL SOURCE RUNS RANGE AVERAGE cm? /cm REMARKS

Niobium 2 + 14.1 to + 24.5 + 19.3 0.2 Continuous film

Molybdenum 2 + 9.9 to + 11.0 + 10.4 0.6 Cont inuous uniform film plus fairly
continuous intermetallic

Vanad ium KAPL 1 + 9.9} 0.4 Heavy continuous film formed some
intergranular attack

Tungsten 2 - 39.6 to + 3.7 + 3.7 0.6 Uniform attack

1040 Steel SAE Steel Co. 3 - 0.3 to + 4.6 + 2:6 0.6 Uniform attack with selected corrosion
of carbide

Tanta lum 2 + 2 to - 2.8 - 2.4 e.6 Uniform film ever original su: face
plus penetration Jlayer

Fe (Armco) Corzy Steel Co. 2 0 to + 6.2 : 0.1 0:6 Uniform attack

1,-605 Haynes Stellite Ceo- 3 - 4 to 4 0.1 - 3.5 1.3 Cont inuocus film intergranular type
attack

Besyllium 3 - 25-4 to - 14,8 = 21 0.6 Fairly uniform attack

Cobalit Kulite Tungsten Co-. 2 - 30 to - 18.4 - 24 0.6 Severe penetration type of attack

Chr omium Univ. of Cincinnati 1 - 126 0.7 Severe uniform attack

Silicon 3 Some weight loss noted

Manganese Electro Manganese Corp. 3 Thin strip tested - partially dissolved

Nickel International Nickel 3 Essentially dissolved

Zirconium Bureau of Mines 3 Essentially dissolved

Titanium Remington Arms 3 Dissolved

Thor ium IsC f,3 Dissolved

Uranium 13 Dissolved

 

 

 

 

 

 

* This value

is probably not due to corrosion.

64
evacuated Armco iron capsules for four hours at 1000°C. Samples were run in
triplicate. The lithium was melted in the protective atmosphere of a helium-
filled dry box, drossed to remove floating impurities, and cast into inverted
hollow capsule plugs, one of which is shown in Fig. 15. After cooling, the
plugs were placed in a desiccator before removing from the dry box. 1In runs
using bismuth, the dry box was not used but loading was carried out by charg-
ing a small ingot of bismuth, previously cast to fit into the capsules. To
prepare for a run, the sample was placed in a capsule along with the coolant
metal and plug, the plug was cold pressed into the capsule, the capsule
evacuated, and the seam between plug and capsule heli-arc welded. The weld
was helium leak tested and the capsule and contents heated to the melting
point of the coolant with a vacuum still applied. The portion of the plug ex-
tending above the capsule was then crimped twice, sheared off by a third crimp,
and spot welded. Heat for the test was furnished by four tilting Burrell
Globar-tube furnaces, each capable of holding two samples. The tilting feature
allows inversion of the capsules after each run, to drain the coolant metal
from the sample. The furnaces were continuously flushed with helium to prevent
oxidation of the iron capsule and to help prevent fire in case of container
failure. After completion of a heating cycle the capsules were placed in a
lathe, grooves cut through the capsule walls, and the plugs removed. This
method 1is preferred to sawing since 1t results in less contamination of coolant
due to chips from the capsule. Lithium or bismuth remaining was removed and
the samples were weighed. The metal samples were sectioned and mounted for
microscopic study, and surface films, 1f present, were identified by X-ray
diffraction. The lithium and bismuth were analyzed spectrographically for

sample components.

The bismuth used was specified as 99.95% pure, supplied by Belmont.
Lithium was obtained from the Maywood Chemical Works of Maywood, New Jersey.
The manufacturer’s specifications indicated a product of 99.5% purity, the
remainder being 0.20 calcium, 0.07 heavy metals, 0.03 iron and aluminum, 0.02
sodium and 0.009 silicon. Subsequently, lithium with 0.005% sodium, as supplied

by the Metalloy Corporation of Minneapolis, Minnesota, was substituted.

The data for the various four hour runs at 1000°C are presented in Tables
4 and 5. They represent, with a few exceptions, weighted values for three
specimens. The results are given in terms of weight change per unit of surface

area rather than as atime rate of weight change, because the corrosion rate 1is

65
UNCLASSIFIED

Fig. 15
Plug
Crimped
B
relgre Welded Spot Welded

.-r":J ‘;.f __'-:pe‘:[rrpr

  

Capsule

Capsule
Assembly

 
 

fl-rl'lllhn Mre bampors by | ey

Exploded View of
Copsule Ports

Liquid Metal Corrosion Test Capsule
Y-1557

66
not likely to be linear, and at present data is available for one time interval
only. Samples for test were made up of flats 3/4 in. X 3/4 in. X 1 in. when
ever possible. Since the volume of coolant material placed in each capsule
was uniform, variation in sample surface is reflected in the column headed
"Ratio-Specimen Area to Coolant Volume.'" The value of this ratio for astandard

specimen 1is 0.6.

Corrosion samples from each of the various elements and alloys tested are
being studied to determine the nature and depth of attack, thickness of film

formation and selective leaching of impurities.

Transverse sections were prepared and mounted adjacent to steel backing
plates used to prevent rounding of the specimen edge during polishing. All
specimens were carefully polished and examined microscopically in the unetched
condition. Etching, 1t was felt, would perhaps destroy films and other

corrosion effects at the liquid-metal interface.

Pure Metals in 1000°C (1832°F) Lithium. Zirconium (graphite melted)
(Mag. 1000X) shows good resistance to attack. A thin continuous adherent film,
identified by X-ray diffraction as ZrN, was found. Film thicknesses were approxi-

mately 0.1 mil and 0.4 mil respectively for two lots of zirconium metal tested.

Cobalt (sintered) (Mag. 500X) shows poor resistance to attack. Average
corrosion depth is one mil. Several deep stringers were noted running to a
depth of 6 mils. The intergranular attack is typical of that due to impurities

at the grain boundaries.

Tantalum (Mag. 1000X) shows good resistance to 1000°C lithium, with for-

mation of continuous thin TaC film.

Armco Iron (Mag. 1000X) shows extremely good resistance to attack. Edge
roughness is presumably due to machining during preparation. All specimens

are now being finished on 600 grade metallographic paper.

Columbium (Niobium) (Mag. 1000X) shows good resistance.to attack. The
gain in weight is attributed to a uniform film formation, approximately 0.1

mil deep. The coating has not yet been identified.

Molybdenum (Mag. 1000X) shows good resistance to short time exposure. The
weight gain is attributed to irregular build-up of Mo,C and MoC.

67
Tungsten (Wolfram) (Mag. 1000X) shows good resistance to lithium attack.

Nickel (electrolytic) (Mag. 100X) shows severe intergranular attack and

deep penetration.

Alloy L605 (Mag. 1000X) shows poor resistance to attack. Selective leach-
ing of one or more impurities is apparent from the zone of voids immediately
below the surface layer. Depth of the selective type of corrosion was approxi-

mately 0.7 mil.

Titanium (Mag. 500X) shows fair resistance. The thin film formed on the
outer surface, 0.1 mil thick, was identified as TiN. The heavier film under-
neath (~ 1 mil thick) and penetrating into the base metal, is believed to be

TiC. This was Remington Arms graphite-melted grade Ti.

Beryllium (Extruded) (Mag. 500X) shows a scattered non-continuous film

formation and large voids. Attack was severe.

Vanadium (Mag. 500X) shows a weight loss and a continuous non-uniform,
unidentified film. Intercrystalline attack, commencing at the specimen sur-

face and running to a depth of approximately 6 mils, was noted.

1040 Steel (Mag. 100X) showed decarburization throughout the entire sample.

Apparently all the carbon has been removed from the sample.

Figures 16, 17, 18, and 19 show micrographs of the various metals tested
in 1000°C (1832°F) lithium in the unetched condition.

Pure Metals in 1000°C (1832°F) Bismuth. Tantalum (Mag. 1000X) shows con-
tinuous but unidentified uniform corrosion product build-up on the surface.
Depth of the product was approximately 0.4 mil. The product is composed of a
heavier film on the outside specimen surface, approximately 0.3 mil thick, and

a diffusion layer penetrating approximately 0.1 mil into the specimen.

Beryllium (Mag. 500X) shows a moderate to heavy attack, of the straight

solution type.

Chromium (Mag. 500X) shows a severe attack, with pure bismuth adhering to

the corroded surface.

Molybdenum (Mag. 1000X) shows a continuous heavy film formation which has
been identified as alpha iron. This presumably results from mass transfer of
iron from the container wall. A thin unidentified layer exists between the

molybdenum specimen and the iron coating.

68
UNCLASSIFIED
CORROSION SAMPLES TESTED IN MOLTEN LITHIUM Fig. 16
I000°C FOR FOUR HOURS |
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH
RN 0 Hepn NIOBIUM _(COLUMBIUM) _ +0.18mg/m2

   
   

 

 

g£¥?éx';%flfgu 1fl§fl
PHOTO NO. Y-1302 000X
WOLFRAM (TUNGSTEN) '1”6«‘##?

 

PHOTO NO. Y1300 T 1000X PHOTO NO.  Y-130I S SN 000N"
UNETCHED

 
UNCLASSIFIED

CORROSION SAMPLES TESTED IN MOLTEN LITHIUM Fig. 17
I000°C FOR FOUR HOURS
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH

ELECTROLYTIC
NICKEL

_ ~90. 95mg/em? ALLOY L605 -7.37mg/om?

 

 

   

 

 

 

 

 

 

PHOTO NO. Y-1296 100X PHOTO NO. Y-I1519 1000X

2
TITANIUM 4SO

 

Lk

UNETCHED B

 

 

 

 

PHOTO NO. Y-ISI7 500X

 
14

UNCLASSIFIED

CORROSION SAMPLES TESTED IN MOLTEN LITHIUM e

I000°C FOR FOUR HOURS
SURFACE AT TOP OF PHOTOMICROGRAPH

 

 

 

 

 

 

 

 

     

EXPOSED
BUREAU OF BURE OF
MINES ZIRCONIUM +0.52mg/cm? MiNES zancomum +0.09mg/cm?
FILM
Fl%m ZrN
ool o Vel o
¢ ' N e s
: Syt AT e ;“;-."f x noaes 2 Vo b LAy &
> e bt R ey L e 1 2t SAR% Sy A AN P
%“ e,*mgfiftw?-"’- BT B AR e SRR LSt - AN I S
PHOTO NO. Y-1295 1000X PHOTO NO. Y-1294 000X
2
SINTERED COBALT . - 4.82mg/cm TANTALUM —I.52mg/cm.’
FILM
) TaC

 

 

1000X

 

 

PHOTO NO. Y-1298

 

UNETCHED
 

CORROSION SAMPLES TESTED IN MOLTEN LITHIUM N aa
I000°C FOR FOUR HOURS
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH

STEEL 1040 -I.BZMM.'. _ —IZZ‘I-I'I'Igfun?

  

 

 

 

 

 

 

el

UNETCHED
R.J. GRAY

 
Armco Iron (Mag. 1000X) shows a good resistance to attack with no apparent
weight change or damage to metal specimens for short-time exposure, but this

may be due to the fact that Armco iron was the container.

Vanadium (Mag. 500X) shows formation of a continuous heavy film (1.8 mils)
identified by X-ray diffraction as alpha iron. It also appears that some

selective corrosion occurred at the grain boundary.

Columbium (Niobium) (Mag. 1000X) shows a continuous unidentified film
formation (approximately 0.1 mil thick) with pure bismuth metal adhering to

the specimen surface.

1040 Steel (Mag. 1000X) shows selective corrosion (decarburization)
occurring near the surface of the sample and extending to a depth of approxi-

mately one mil.

Sintered Cobalt (Mag. 500X) shows drastic attack, penetrating deeply into
the specimen. Apparently the attack proceeded down through the sintered
powdered particles and left a matrix of unattacked metal behind. Pure bismuth

is present in the network.

Tungsten (Wolfram) (Mag. 1000X) shows a heavy uniform straight solution
type of attack.

Alloy L605 (Mag. 500X) shows severe intercrystalline attack to a depth of
approximately 6 mils. A continuous film was formed on the specimen surface,

approximately 1 mil in thickness, identified as gamma 1iron.

Figures 20, 21 and 22 show micrographs of the various metals tested 1in
1000°C (1832°F) bismuth in the unetched condition.

The static tests are as yet quite incomplete, but even from this fragmen-
tary information some materials can be eliminated from consideration. Others
can be selected as having sufficient promise to justify further study and con-
sideration, but the available information does not by any means warrant their

unqualified recommendation as reactor materials.

In the case of bismuth the materials which show promise areirun, molybde-
num, tungsten, and tantalum. The same materials, together with zirconium and

a number of iron-base stainless-type alloys, warrant further study in lithium.

These and other tests not yet reported indicate the importance of giving
consideration to the presence and nature of any third component which may be

present in the system. Mass transfer of relatively insoluble materials—so-

73
CORROSION SAMPLES TESTED IN MOLTEN BISMUTH G i
I000°C FOR FOUR HOURS |
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH

SINTERED COBALT -2997mg./cm? WOLFRAM (TUNGSTEN) +3.69mg/cm?

 

    
       

 

 

 

 

 

 

B nns st WO e By
1% S § gl N eneed Toadin B ;
PRV Y Ty S PHOTO NO. Y-1409 500X

ALLOY L605 - 3.69mg/cm®

 

UNETCHED R.J. GRAY

 

 

 

 
CORROSION SAMPLES TESTED IN MOLTEN BISMUTH S
I000°C FOR FOUR HOURS
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH

ARMCO IRON +0.18mg/cm? VANADIUM +9.98mg/cm?2

 

 

 

 

 

 

 

 

»® .
B
PHOTO NO. Y-14! 1000X PHOTO NO. Y-1416 | . 500X
NIOBIUM  (COLUMBIUM) +24.3Img./cm? STEEL 1040 -0.33mg/m?

   
   

     

s
A i
& L

¥
PO W0 Y-1388 000X PHOTO NO. Y-Mi2
UNETCHED

 

 

 

 

 

1000X
R.J. GRAY
UNCLASSIFIED

CORROSION SAMPLES TESTED IN MOLTEN BISMUTH Fig. 22
I000°C FOR FOUR HOURS
EXPOSED SURFACE AT TOP OF PHOTOMICROGRAPH

TANTALUM =l  +2.95mg./om? EXT. BERYLLIUM ~14.84mg/cm?

 

    

 

 

 

 

 

 

N
a ' \- A
PHOTO NO. Y-1389 000X PHOTO NO Y-H410 000X
ELECT.CHROMIUM —I126.40mg/cm? MOLYBDENUM +9.90mg./cm?

%
P gl el oy o s PR W e Ty, et s ey 4

 

 

 

 

 

 

PHOTO NO Y-1390 000X

 

UNETCHED
R.J. GRAY
lution from one location and precipitation at another location, probably on
the third component, without at any time being present in high concentrations

in the liquid--appears to be a serious factor.

Dissolved Fuel Tests. Because of the recent interest in the liquid metal
fuel type reactor, some exploratory corrosion tests were conducted to find
potential materials for containing uranium bearing coolants. The lead-uranium
and bismuth-uranium binary systems look somewhat promising froma uranium solu-
bility viewpoint and for this reason were selected for these tests, The liter-
ature shows lead will dissolveup to 0.5 percent uraniumat 900°C (1652°F) while
upwards of 25 atomic percent uranium is soluble in liquid bismuthat 1000°C
(1832°F). The minimum solubility of uranium in the coolant thought necessary
for successful reactor operation is of the order of one atomic percent.

The materials tested were confined to some of the high melting metallic

elements:
Tungsten Titanium
Tantalum Iron
Molybdenum Nickel
Columbium Beryllium

Zirconium

Corrosion specimens 1 in. X 1/4 in. X 3 in. were prepared from strip
stock and were immersed in the uranium bearing coolant by means of a vertical
positioning rod. (See sectional view of test apparatus in Fig. 23.) The BeO
crucible containing the coolant was housed in a cylindrical guartz tube heated
externally by an electric resistance-type furnace. Tube ends were sealed with
metal and rubber gasket fittings to permit conventional evacuation and purging

of the test chamber prior to operation under an inert atmosphere.

The bismuth and lead bath metals contained approximately two atomic per-
cent uranium for all tests. This amount may be in excess of the solubility of
uranium in lead, however. Test temperature was 1000°C and test duration was

four hours at temperature.

Preliminary results showed substantial attack on most materials with for-
mation of films in most cases. Beryllium in the lead-uranium bath formed a
UBe, coating, and also formed a coating (X-ray identification not completed) in
bismuth-uranium. Molybdenum and tungsten showed the least attack of materials

tested, but the results should be confirmed in longer tests which are being

made .

17
ARGON QUTLET

WATER COOLED Sl -

TUBE-END FITTING

 

J

 

 

 

 

 

 

 

 

 

2

NOT CLASSIFIED
DWG. 9082

Fig. 23

VACUUM SHAFT SEAL

POSITIONING ROD

REFACTORY SHIELD

 

= 1Y //@

 

QUARTZ TUBE

 

 

b
X

 

 

 

i L ¥ X ¥ X ¥

 

 

 

 

 

 

 

 

  

PEDESTAL
BLOCK

    

 

    
 
  

  

EATING
ELEMENTS

 

 

 

l T— = Al ]
ol hc>/)<' /))<;;X/
|Of tfe
101 1 0| /CORROSION
ol ~ Hol SPECIMEN
O- P'O
Ol 10
é;i BeO Iig
of CRUCIBLE o
O 1O
0| | O
O
O
O
oy

 

 

ARGON INLET

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S
S
\\

VACUUM OUTLET

_— RUBBER GASKET

SECTIONAL VIEW VACUUM ARGON CORROSION TESTING APPARATUS

78
DYNAMIC CORROSION TESTS

Thermal convection Loops (Harps). Twelve convection loops are being made
by the Philadelphia Pipe Bending Company. Six of the loops will be of 310
stainless steel, three of V-36 alloy, and three of L-605 alloy. Three of the
310 loops and one each of the L-605 and V-36 loops will be constructed of
material shrouded over a tube of 1010 steel. The fabricator has experienced
considerable trouble in trying to weld the sections of shrouded material to-
gether and still maintain an unbroken inner surface of iron. At present the
base metal is 0.080 in. thick and the inner shell is 0.020 in. thick. One of
the 310 stainless steel loops fabricated at Philadelphia will be sent to
Battelle Memorial Institute for internal cladding with pure iron by vapor

deposition.’

" The OBNL machine shop has fabricated one loop of a low carbon, deep-draw-
ing, fully-killed steel and is fabricating two more of this low carbon steel
and four each of 304 and 347 stainless steels. Material has been ordered for
446 stainless steel and for nickel loops, and efforts are being made to obtain

loops of zirconium and molybdenum.

The engineering design group has drawn up a specification for a cylindri-
cal pressure vessel to serve as a dry box for the thermal convection loops.:
The specification has been sent out for bids for three of the pressure vessels.
The auxiliary equipment for the pressure vessels, such as heater and thermo-
couple cap assemblies, mounting and terminal plate assemblies, and air cooling
jackets, are either on order or have been designed and are ready to be turned

over to the shop for fabrication.

A lithium purifier constructed of 316 stainless steel with a capacity of
ten pounds of lithium has been designed  The 316 stainless steel plate tubing
and piping have been ordered and delivery of the material is expected within

two weeks.

A thermal convection loop of the very low carbon steel has been filled
with a lead bismuth alloy containing 47% bismuth. This loop will be operated
at temperatures up to 800°F. As soon as a loop of high temperature oxidation
resistant material is obtained, a similar setup will be made to operate at
temperatures up to 1800°F, The heating characteristics and power requirements
of the various parts of the loop will be determined, and the control and

electrical equipment needed teo operate the harps will then be determined.:

79
Forced Circulation System. A system is being planned to serve as a simu-
lated service corrosion test for reactor and heat transfer components. It is
intended that this system will duplicate, as far as possible, the anticipated
materials, component construction, velocities, and temperatures of the proposed
Aircraft Reactor Experiment, although it will not duplicate the radiation

effects or the total number of components planned for the reactor.:

80
Sy i

LITHIUM ISOTOPE SEPARATION

G. H. Clewett

Chemical Research Division

The very superior heat transfer properties of lithium make it look very
attractive as an aircraft reactor coolant. However, the normal 7.35 percent

Li® content must be largely eliminated if the liquid is not to have too great

‘& neutron absorption cross section; therefore, all of the likely techniques

for producing tonnage quantities of separated Li’ isotope, in a purity about

99.93% at a reasonable price, are now belng explored / The various phys1co~

s e e ety

chemical methods are discussed in this section.

The small weight of the lithium isotopes with resultant large percentage
difference in mass would seem to make the problem of separation of these
isotopes arelatively easy matter i1f purely physical methods are used. Because
of this a considerable portion of the total effort on this problem has been
placed on development of the molecular distillation method which now appears
to be entirely feasible, but must undergo considerable engineering development.
It is further apparent that the very simplicity of the chemistry of lithium
with its single valence state in compounds, the tendency of the 1on toward
extreme solvation in solution, and the dearth of complex forms of the element
would appear to make the search for chemical methods somewhat difficult. This
is true; however, the ease of handling a chemical process, possibly a two
phase liquid-liquid system, in conventional equipment at ordinary temperatures
and the fact that it operates at, or close to, thermodynamic equilibrium at
each stage appears to be sufficient justification for putting a sizable
portion of the effort into the search for such a chemical system. Two phase
liquid-solid systems must also be studied with perhaps less vigor but with
some care and thought since sizable improvements continue to be made in

methods of engineering liquid-solid contacting processes.
MOLECULAR DISTILLATION

Definite progress has been made toward the separation of lithium isotopes

1

i
|

by distillation methods. A positive separation factor has been established |

and a tentative production rate determined. Several additional experiments in

81
a single stage still (Fig. 24) have established a separation factor a = 1.02
or slightly greater. As a result of distilliation over a fairly narrow temper-
ature range (at around 450°C) some indication of evaporation rate, and there-

fore production rate, has been determined.

The experiments with the one stage still pointed to several changes and
improvements in design that would be essential to smooth operation, As a
result of this experience a two stage still (Figs. 25 and 26) was designed
and fabricated which incorporated these modifications. This still has pro-

vision for reflux of liquid and the condenser-roof is so designed that the

vapor progresses from stage to stage. Considerable time was spent in testing

this apparatus and it has been only recently that sufficiently satisfactory

operation was achieved to warrant sampling the product for assay.

A further step in the evaluation of the molecular distillation method
has been the design and construction of a simple evaporation apparatus. This
unit consists of a single evaporating surface and condenser and is to be used
to determine the rate of evaporation of liquid lithium at various temperatures
and pressures. The equipment has been test run only. Data which will be
secured from this equipment should prove very useful in engineering calculations

on multi-stage still design.

A packed column refluxing still has been designed and fabricated. This
will be used to test the feasibility of higher temperature and higher pressure

distillations of lithium, It should be placed in operation in the near future.

Considerable thought and effort have been devoted to the design of a
multi-stage vertical molecular still, It.is being held in abeyance until more

information is available from the smaller equipment now being operated.

CHEMICAL EXCHANGE IN A LIQUID-LIQUID SYSTEM

Because of the many advantages of a liquid-liquid countercurrent process
over a liquid-solid countercurrent process the search for such a system has
been rather painstaking and thorough during this quarter. Furthermore, since
one of the methods reported in the published literature, using lithium amalgam-
lithium salt solution, was of this category, it appeared that some of the

development work should be in this direction.

82
SCHEMATIC

 

 

   

 

 

 

 

 

 

 

 

D
o
v

 

 

 

 

v’/}j////// A
® U
KEY
() ROOF (7)) ROOF SUPPORT
® BOX ~__—(® RECEIVER
(3) STILLPOT (@ BOTTOM PLATE
@) INSULATION GAUGE CONNECT.
®) SPARK PLUGS @ VACUUM LINE
® HEATER (2 BONNET

(® WATER LINES FOR BOTTOM PLATE

 

Wy
v /
..

s -l'-..

FIGURE 24 SINGLE STAGE LITHIUM STILL

@3IdISSYTIONN

€8
 

 

- e

 

1.
2.

4.
5.
6.
T
8.
9.
10.
11.
12.
13.
14.
15.
16.

 

Key for Fig. 25

Vacuum gage connection

Vacuum bonnet

Thermocouple wells

Radiation baffles

Coolant ducts

Still pot assembly

C;ndanser assembly

Heating elements

Flexible connection for coolant
Radiation baffle

Reflux ducts

Vacuum gasket

Spark plugs for electrical leads
Still support

Wilson vacuum seal

External coolant connection l

84

 

 

 
NOT CLASSIFIED FIG. 25

2 s meL) Dwg. 9232

s BN A

 

=
A A A

 

 

 

 

 

 

.

IAEIIIIHIRITHIIIIRIANRY

\‘&\\\\\‘m\fi\\\ -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

s A R,
\'*--H.__q ;——--"_HZ @
7
@\/ b é 7
é ?
@—p e =T ff
? NSl | LA ?
g ||| (I e S
g L NN N AN AR GNGN . .r- é
_ /

 
 

 

  
  

 

 

 

 

/ 2

TO VACUUM SYSTEM

 

WO STAGE LITHIUM STILL

[
_g5— /7 NOT CLASSIFIED
B R e

Y-12 PHOTO 6-1460
/ NOT CLASSIFIED

L] '!'-'-‘f_:'.ili,lihifl"-

ST

TP -, 7 J
. - N f o - o bl LT e

_ Gl Wy =I= . ' e ',!I*a.__': |
B | TWO STAGE LITHIUM STILL

4

 
Lithium Amalgam-Lithium Salt Solution. Considerable effort has been
devoted to a study of the lithium amalgam-organic solution exchange as first
reported by Lewis and McDonald.(24) Initial experiments using ethanol (95%) made
it very apparent that the rapid reaction of the water and alcohol with the
lithium in the amalgam would make this system unsuited for sustained operation
unless this reaction could be either eliminated or materially reduced. Experi-
ments with absolute alcohol (H,0 about 700 ppm) indicated that although the
moisture content was a very important contributor to the reaction, the alcohol
itself reacted too rapidly to be practicable, This discovery led to an in-
vestigation of the higher alcohols up to isoamyl. It was found that the rate
of reaction decreased substantially as the carbon chain was lengthened and it
appeared from static tests on hydrogen evolution that isoamyl alcohol had
definite possibilities. It soon became apparent that a new difficulty must
be overcome, namely emulsions. Isoamyl alcohol, with or without dissolved
lithium chloride, forms very stable emulsions with lithium amalgam. More
than 150 organic solvents of suitable structure were examined for their ability
to stabilize the system against the formation of emulsion. No success was
achieved with this program. It was discovered, however, that those materials
which were known to react chemically with the lithium in the amalgam were very
effective in breaking and preventing emulsions. In this group are water,

methanol, ethanol, etc.

Following this initial disappointment with the amalgam system both from
reaction and emulsion standpoint, a survey was made to find some organic
solvent which would neither react chemically nor cause emulsions. Of a rather
wide group of organic liquids representing nearly all the major classes of
compounds not a single example has been found which measures up to the de-
sired characteristics. As a matter of fact it appears that the following
generalization can be made: those organic liquids which do not have a re-
placeable hydrogen (or hydroxyl group) atomwhen contacted with lithium amalgam
form stable emulsion with the amalgam, All indications point to the con-
clusion that the lithium is selectively located at the emulsion interface as

the stabilizing agent for the dispersion.

The magnitude of the separation factor of the amalgam exchange system

has been determined by batch shake-out methods. The procedure was to contact

(24) Lewis, G. N. and MacDonald, R. I. "The Separation of Lithium Isotopes," Am. Chem. J. 58, 2519-2524,

(1936).

87

1
'
lithium amalgam (about 0.5 molar) with an absolute ethanol (700 ppm moisture)
solution of lithium chloride for two minutes. The amalgam was then divided
in two equal parts and a new alcohol solution made from the lithium in one

of the portions. These operations were carried through five stages.

Reaction of the lithium in the amalgam with the alcohol solutions caused
a 33% depletion during the five operations. The over-all separation effect
observed from this experiment shows o to be 1.0495 as based on a single set of
assay results. This high separation factor makes it almost mandatory to
continue the search for a suitable method of utilizing the amalgam system for

separation of the lithium isotopes.,

Aqueous-0rganic Systems., -Although an aqueous phase in contact with an
immiscible organic phase is one of the first possibilities which suggested
itself for a two phase system, it now appears that only selected examples of
this type or greatly modified solutions suchas lightly salted aqueous solutions
may approach the arbitrary criteria set up for suitable exchange systems.
Since the distribution of lithium chloride between water and isoamyl alcohol
is not completely unfavorable, two separate batch extraction experiments were
carried out on this system., It now appears, after a recheck on the assay of
the first experiment that no significant enrichment was obtained in these two
attempts. A somewhat more favorable distribution ratio of lithium salt be-
tween phases can be obtained 1f the aqueous phase of any two phase system 1s
highly salted with some inert salt. This phenomenon is being studied at some
length and multiple batch separation experiments will ptobably be conducted

on selected systems of this type.

Organic-0Organic Systems. The predominantly hydrophillic properties of

lithium ions prompted an investigation of the possibility of discovering a

‘suitable pair of immiscible organic solvents as the liquid-liquid media for

a separation system. More than 400 organic-organic systems were examined.
Of these, approximately 60 were found to be immiscible pairs. Many of these
pairs were rejected from qualitative observations of viscosity and phase
separation. Of the remaining systems 12 pairs show sufficient promise to
warrant more thorough investigation. These pairs are:

1. Butyl acetate vs., ethylene glycol

2. Hexone vs. ethylene glycol

3. Benzaldehyde vs. ethylene glycol

88
- Dimethyl aniline vs. ethylene glycol

. Ethyl benzoate vs. ethylene glycol

4
S5
6. Nitromethane vs. octyl alcohol
7. Formamide vs. butyl acetate

8

. Formamide vs. hexone
9. Formamide vs, octyl alcohol
10. Formamide vs. ethyl ether
11. Formamide vs. dimethyl aniline

12. Formamide vs. ethyl benzoate

These:systems are now in process of investigation to determine the
distribution of lithium in equal volumes of the liquids and the saturation

solubilities of various lithium salts.

An added phase of this study has been the preparation of special lithium
compounds which would appear to possess the desired solubility properties for
for these organic solvents., Among those already prepared in this laboratory

have been the following:

1. Lithium perfluorobutyrate (LiC4F702)
2. Lithium thiocyanate (in hexone)
3. Lithium salicylate

(The work on these varied projects is being intensified.)

Pulse Column Application. Pulse columns(zs) have been investigated for
their possible application to the lithium amalgam system. These columns have
several advantages over ordinary countercurrent gravity columns, the principle

one being short stage length.

The pulse column work was carried out concurrently with the laboratory
examination of the amalgam system and, if anything, highlighted the diffi-

culties inherent in the reaction and emulsion problems.

A series of experiments were made to test the effective height of a
theoretical stage. These were made by contacting a lithium amalgam with a
solution of sodium ions. The apparatus used is shown in Figs. 27 and 28.
The results were inconclusive because of the fact that reaction with the
solvent medium in all cases nearly depleted the alkali metal content of the
amalgam.,

? (25) /A description of the pulse column is given inm U. S, Patent 2,011,186 (1935): HW 14728, The Design
J and Operation of the Pulse Column, Burns, Groot and Slansky (Oct. 12, 1949).

89
I I'Ad L1

DWG, ‘.?Zi

 

 

 

 

 

 

 

 

 

 

 

 

 

 

' i Vent
s »
- e
——
B — Extractant
2 '- Reservo/r
~For Detall see
-~V  Figure 28
o] Furmp

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Samp/1ng
ARGt sy Ml e =
O e G
= e
: | Pulse Froducer
1
T oL

 

 

A mialgars
Reservoir

 

 

 

Fump

PULSE COLUMN

90
1Y, O
DWG. 9230

 

 

71" Glass tubing

spdcer.

" Standard Fyrex Fipe.

 

 

 

Wire Screen—<_ it

 

PULSE COLUMN

Ql
A series of tests were made to determine the relationship between screen
size and flooding rate in the one-inch glass pulse column. This data 1is
summarized in Table 6 which shows flooding rates observed in the system mercury
vs, 95% ethanol which. is believed to be a close approximation to any amalgam

system which might be used.

 

 

 

TABLE 6
SIZE. OF SCREENS (1l.imn. Separatiocns) RATE OF FLOODING
MESH - WIRE HOLE (ce/min of Mercury)
s (in.) (iaio)
40 010 015 >240
50 .009 011 >218
60 . 0075 . 0092 >227
70 . 0065 ., 0078 >242
:80 . 0055 0070 >247
90 .00525 . 0058 26.7
100 50035 - 0055 19.9

 

 

 

 

 

 

Several runs were made to test the action of amalgam against lithium
chloride solutions in the pulse column. With the lower alcohol solvents the
reaction rate proved excessive. In one experiment using absolute alcohkol
(700 ppm moisture) the initial rate of loss of lithium from the amalgam
(0.5 molar in lithium) was 2.8% loss per minute and an over-all loss of 96%
was observed in 47 minutes in this experiment. Special efforts were made
prior to this test to insure that the equipment was absolutely dry and the

entire test was made under inerft atmosphere to reduce moisture pickup.

The emulsifying tendency of such solvents as n-butyl alcohol, isoamyl
alcohol, and pyridine was demonstrated to be a major problem by operation in
1the pulse column, and the selective removal of the lithium from the bulk of
‘the mercury to the dispersed phase was clearly demonstrated in several of

chese tests.

92
The pulse column is now being used to make preliminary investigations on
some of the more promising organic-organic systems that are under investigation

in this laboratory.

CHEMICAL EXCHANGE IN A LIQUID-SOLID SYSTEM

The practical application of a countercurrent liquid-solid system is

considerably more difficult than either the liquid-liquid or liquid-gas
systems. In spite of this fact, some of the more promising liquid-solid

systems have been investigated.

A cellulose column was set up using filter aid pads as the packing
material. Five separate rumns were made using various eluents to remove the
lithium ions previously deposited at the head of the column. The results of
these tests were sufficiently discouraging to cause abandonment of this pro-

ject.

Another liquid-solid system tried comprised a column filled with finely
ground lithium carbonate and eluted slowly with water. As was the case in the
cellulose experiment, no enrichment of the isotopes was observed and this

project was also discontinued.

Isotope separation by the use of synthetic zeolites, such as Dowex 50,
is being investigated in the X-10 area of ORNL. Some success has been achieved
by use of this technique. This work is reported in detail in the Chemistry

(26)

Division quarterly reports.

AUXILIARY STUDIES

All the lithium assays made to date have been determined by use of a

modified Nier mass spectrometer. Another method of assay is being developed

using fission counting technique in which the a particles from the °1i neutron

reaction are counted. Although still in the development state, this method

appears promising,

Several attempts have been made to determine the vibration frequency of

‘the lithium ion-solvent bond in the Raman range. No distinct band which could

(26) Swartout, J. A., Chemistry Division Quarterly Progress Report for Period Ending March 31, 1950,
Part I (June 16, 1950).

93
I be attributed to the lithium-solvent bond was observed. This work has been

abandoned, at least temporarily.

o1 g

94
RADIATION DAMAGE

REACTOR MATERIALS
D. S. Billington, Metallurgy Division

The behavior of materials when exposed to the high neutron flux antic-
ipated in an aircraft reactor is of extreme importance in the design of such
a reactor since it must be assumed that material properties of importance may
change significantly.  There is no adequate theory which can give useful pre-
dictions and experimental data are at present meager. Experiments in this
field are difficult and often very lengthy, soit is important that experiments
which will answer the unknowns involved in the construction of the reactor
be initiated in the near future.: Planning such experiments is complicated by
the present uncertainties about the reactor itself, as neither materials to
be used nor exact design have been determined.: A further complication is that
neutron fluxes of the proper magnitude do not exist., However, experiments are
in progress and more are planned making use of both accelerators and piles,
which will yield information of particular interest for reactors as envisaged
at present as well as information of a general nature. The equipment and
experience resulting from these experiments can be applied to different
materials and design problems as the requirements become known. Fuel element
design has not yet progressed to the point where testing of samples can be

started profitably.

Accelerator Experiments.

(a) An experiment has been set up by North American Aviation, Inc. that
makes it possible to study the effect of deuteron bombardment on the corrosion
rate of Armco iron by lithium at 1000°C. It is thought the deuteron bombard-
ment will simulate to a high degree the behavior of this system under neutron
bombardment, and thus will give answers on this important problemina relatively
short time. The apparatus has been assembled and is being bench-tested at the
present time.  Irradiation data are expected very shortly. It is planned to
extend the experiment to include molybdenum in contact with both lithium and
sodium, also at 1000°C.- An ORNL man has joined the NAA group for the purpose

of assisting in the experiments.:

95
(b) The Purdue Group is expected to begin work by July 1. The following

experiments are tentatively being planned:

(1) Study of SiC as an analog material for Be,C.* Measurement of
electrical property change induced by deuteron bombardment.

(2) Effect of bombardment on creep rate.

(2a) Repetition of Andrade experiment(?’J) using alpha particles
and deuterons. '

(2b) Effect of bombardment on creep rate of stainless steel and
other high temperature materials

(2¢) Effect of bombardment on creep rate of single crystals akove
and below recrystallization temperature.

In-Pile Creep Tests. A creep test apparatus (illustrated schematically
in Fig. 29) capable of operation under irradiation in a stringer in the ORNL
pile is being developed. In the preliminary model now being bench-tested, a
specimen nine inches long and one-eighth inch in diameter is enclosed 1in a
constant-temperature furnace throughout about two-thirds of its length. The
specimen is to be stressed by means of cables extending outside the pile where
a dead weight furnishes a constant load. The central four-inch section of
the specimen is the gauge length over which the strains are measured. Longi-
tudinal strain is transmitted to a microformer by extensometer elements welded
to the ends of the gauge length. A duplicate microformer whose core may be
moved a known distance by an electromagnet controlled from outside the pile
will be included in the apparatus alongside the strain measuring microformer
so that any effect of temperature and radiation may be compensated for. An
entire duplicate creep apparatus may be placed in series with the load supply
cables outside the pile so that the difference in creep rates, in-pile and

out-of-pile, may be observed.

In bench tests the temperature control system now in use has proved
capable of holding the test bar temperature to within * 1° at 1200°F. The
variation of temperature over the gauge length is now % 4°F. Work is now

directed toward minimizing this value to # 1°F, at which time the bar will

be loaded to test the loading system and strain measuring instruments.

* Accelerator experiments on Be, BeO, Be C, and Zr using particles above 1 Kev are now classified. The
possibility of using Mg2Si as an isoelectric analog material for Be2C is being studied, -

(27) Andrade, E. N, da C., "Effect of Alpha-Ray Bombardment on Glide in Metal Single Crystals,' Nature
156, 113 (1945).

96
L

Graphite Mmatrix

Iron Furnace Tibe Nichrome Heater Windings

Graphite Mica /Insulation Quartz
Stringer

Quartz Tube
ensometer Spring

Microformer
G4 E xtensameter Flements
Spot Welded To Test Bar

 

Loading Wires To
Dead Weight At Pile Face

IN PiLe Creep ArPparATUS
for Insertion In Stringer In ORNL Pile

Dwg. No. 9146
Frg. 29

—
The above apparatus is designed to supplement the creep experiments of
the NEPA Group = It is felt that between the two groups all ranges of temper-
atures and types of materials of interest can be studied in the existing pile

facilities.,

The NEPA and ORNL Groups are investigating jointly with Argonne the use
of special facility at Hanford that will permit dead-weight loading of creep
units.  Use of dead-weight loading permits considerable simplification in the

design of creep units.

At this time it appears desirable to study the creep properties of pure
metals such as molybdenum and iron and alloys of the stainless type in addition

to the ceramics being studied by NEPA.

Diffusion of Fission Products. It appears necessary to obtain more data
on the diffusion of fission gases and fission products in various metals than
is presently available in the literature. For example, it seems evident that
the fission gases do diffuse out of undamaged uranium, though at a rate so
small as to be insignificant in times of interest in an aircraft reactor. It
also seems evident that the inert fission gases will not diffuse through
metals such as stainless steel even at 1000°C in the absence of radiation.
This seems to sum up our immediately useful knowledge. Thus 1t 1s necessary
to measure diffusion rates of uranium in molybdenum and other high temperature

materials with and without irradiation.

There seems to be a lack of information also on the diffusion of the
halogens at high temperatures. Nor is it known whether a thin cladding of in-
ert metal on uranium would prevent accumulation of fission gases. In an
attempt to gain an answer to these and other problems diffusion experiments 1in

and out of the reactor are being planned.

Hot laboratory space is being made available, and it is expected that
some measurements such as hardness, electrical resistivity, elastic modulus,
magnetic susceptibility, and tensile strength on some samples of Inconel,
Hastelloy, stainless steel, and zirconjum that were irradiated in the Argonne
channel experiment in the X-10 reactor will be obtained. Measurements will

also be started on some stainless steel samples irradiated at Hanford.-
Samples of Ti, Mo, Ni; and Armco Fe suitable for mechanical property
measurements will go into the "H" pile via the Argonne channel experiment,

according to the latest revised Hanford schedule, on June 29, 1950.

AUXILIARY MATERIALS

O. Sisman, Reactor Technology Division

Plastics. This work has been temporarily delayed because of the con-
struction work which has been going on in the pile building. It 1is expected,
however, that most of the normal pile irradiations will be completed in the
next three months. Work is also planned to determine the effect on radiation
stability due to irradiation rate, oxygen and antioxygens, and to determine
the damage due to thermal neutrons, fast neutrons and gamma radiation. Design

of an apparatus for subjecting specimens to a 10° r/hr gamma source is nearly

complete.

Metal Hydrides. The samples of titanium, zirconium, and lithium hydrides,
on which data were presented in the last quarterly report, have been removed
from the pile and are being held in storage until they have decayed sufficiently
to permit a chemical analysis on the materials and the containers. It is
hoped that a better interpretation of the data will be possible after the

chemical analysis.

Preparations are being made for further studies on titanium hydride,
lithium hydride and uranium hydride. The apparatus for determining the dis-
sociation pressures of these materials under pile radiation up to temperatures

of about 500°C will be installed in the pile during the coming period.:

Liquid Metals In-Pile Experiment. To fully evaluate the feasibility of
using liquid metals as a primary reactor coolant it is desirable to study these
materials in a flowing system under radiation. Work was started in April on
the design of a loop (Fig. 30) in which liquid metals may be circulated through
the X 10 reactor at high temperatures. The primary data to be gained from
such an experiment will be (1) to‘determine the activity pickup of the liquid
metal from the pipe through which it is circulated, (2) to determine the
radiation effect on corrosion (and erosion) rate, and (3) to develop techniques

for handling liquid metals under radiation at high temperatures.:

99
00l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INSULATION ALUMINUM RETAINER 2" 0D TUBE ) RADIATION  * GHARGE
\ \ \ 2 0D COUNTER J LINE
IR A IR RZ ] Tuee N
R0 . ‘i\\ ' VACUUM OR
WWXWXX?& @ -1 L FLOW:. ;EINERT GAS
{ vt { 1
< APPROX 19" IN PILE —————> :: METER | g;/,/:?:.-—->
f ,.--ll--*{, ') ':';
i .'” i || | b
VAGUUM Loty P
I DT
“ r--: | :
\‘\.\ ‘ |
APPROX 20 Y \,\:‘_1-_-__:_-{':_—_-_1 !
;"", _ _ 1
- -d ’\
7 EXPANSION
E.M. PUMP TANK
GE GAT. NO. 9159849G2
[e———— apPROX 3’ o]
NOTE ; DASHED LINES INDICATE TENTATIVE DESIGN
LIQUID METALS "IN-PILE" EXPERIMENT

_—
T€26 "DuQ
o¢ ‘614
The first experiment will use 316 stainless steel as the container and
the liquid metal will be lithium, but the equipment will be designed so that
the loop containing the liquid metal may be readily interchanged and any com-
bination of container and liquid metal may be used. It would be most desirable
to use Li’, instead of natural lithium, and a request has been placed with
Y-12 for anestimate of the time required and the cost to produce two pounds of
Li7 by electromagnetic separation. It is thought that a purity of 99.5% Li’
will be sufficient for this experiment although 99 93% is the estimated require-
ment for the final reactor. In the event that Li’ is not available, a highly
purified natural lithium will be used (the sodium, in particular, must be re-

moved ).

‘Design of all internal portions {(parts that will fitinto the pile proper)
has been completed and turned over to the shop for fabrication. This portion
of the apparatus consists of an electrically heated loop of 5/16 in. OD tubing
appro£imately 20 feet long to contain the lithium. A 2 in. evacuated tube
jackets the loop and contains foil radiation shields. The jacket also serves
as a vacuum chamber which will insulate against heat loss by convection and
at the same time aid in detecting lithium leaks since the lithium vapor will
cause a decided pressure rise in the vacuum system. Insulation will be packéd
around the jacket and will be retained by an aluminum can of 3-3/4 in. % 3-3/4 in.

outside dimensions.

Since the maximum allowable temperature of the outside of the aluminum
can is 250°C in the pile and 75°C in the concrete of the pile shield and since
it is desired to maintain the lithium ultimately at 1000°C, the internal parts
and vacuum system will be assembled first and heat transfer data obtained to

determine whether a cooling coil is needed in the insulation.

The external portion of the system will consist of an electrically heated
reservoir over which an inert atmosphere will be maintained, anelectromagnetic
pump, an electromagnetic flow meter, and activity detecting equipment. An
a-c electromagnetic pump is on order from General Electric, and delivery 1is

expected in June . The flow meter has not yet been designed.

Calculations are being made on the activity to be expected in the loop
based on the pessimistic assumption that half of the 316 tubing will be re-
moved to the lithium From the results of these calculations decisions will
be made about the type of activity counting equipment required and the neces-
sary shielding for the portions of the system which will be outside of the

pile.
101
NUCLEAR MEASUREMENTS

A. H. Snell, Physics Division

It seems quite likely that the eventual aircraft reactor as well as the
ARE will be of the intermediate type. This will create a demand for greatly
improved intermediate cross-section data for several elements. Both experi-

mental and theoretical efforts are underway to accumulate this information.

HIGH VOLTAGE PROGRAM

A High Voltage Laboratory building has been conceptually designed.: It
will house the 5 Mev NEPA Van de Graff machine along with the present 2 Mev
machine and the Cockcroft-Walton set. .- It is hoped that detailed design and
construction can be rushed so that the 5 Mev machine will not wait in idleness
for more than a few months. Two men have been obtained for cross-section
measurements. They are experienced and interested in spectrometry of this

nature. -

MECHANICAL VELOCITY SELECTOR

AEC approval of the change recommendation covering construction of the
mechanical velocity selector is expected momentarily. After considerable
study, a tool steel rotor about 12 inches in diameter has been decided upon.
This will contain hydrogenous scattering material with slits, and will rotate
at about 10,000 rpm. An 80-channel recording system is envisaged.” A second

man will join the project this month

102
PUBLICATIONS

(List of External Reports Issued puring the Last Quarter)

ORNL 535 Determination of the Fast Flux in Hole 19
O. Sisman and C. D. Bopp

ORNL 649 Uranium Hydride, A Survey
A.-S. Kitzes

ORNL 665 The Approach to Critical With the Bare Intermediate Reactor
N. M. Smith

ORNL 668 A Diffusion Solution for the Cylindrical Ducting Problem of

Infinite Geometry
D. Whitcombe

ORNL 684 The Atomic-Powered Aircraft, January 1950
C. B. Ellis
ORNL 710 Theoretical and Practical Aspect of Shzeldzng

A. S. Kitzes and T. BRockwell

103 o