ORNL-2723.txt

i CENTRAL RESEARCH LIBRARY
|| | |||{"|H || II DOCUMENT COLLECTION

3 4456 0361372 O

- ORNL=2723
Reactors=Power

MOLTEN-SALT REACTOR PROJECT
QUARTERLY PROGRESS REPORT
FOR PERIOD ENDING APRIL 30, 1959

CENTRAL RESEARCH LIBRARY

DOCUMENT COLLECTION

LIBRARY LOAN COPY
DO NOT TRANSFER TO ANOTHER PERSON

I you wish someone else to see this
document, send in name with document
and the library will arrange a loan.

OAK RIDGE NATIONAL LABORATORY
operated by

UNION CARBIDE CORPORATION
for the

U.S. ATOMIC ENERGY COMMISSION

/LY
ORNL-2723

Reactors — Power
TID-4500 (1h4th ed.)

Contract No. W-7L05-eng-26

MOLTEN~-SALT REACTOR PROJECT
QUARTERLY PROGRESS REPORT

For Period Ending April 30, 1959

H. G. MacPherson, Project Coordinator

DATE ISSUED

JUN 191953

OAK RIDGE NATIONAI LABORATORY
Oak Ridge, Tennessee
operated by
UNION CARBIDE CORPORATION
for the
U.S. ATOMIC ENERGY COMMISSION

GY SYSTEMS LIBRARIES

(TR

3 yysk 0363378 0

s

CONTENTS

SUWARY -o----.o-aooc-o.-o-ooo-.-eo-oo-on-.o-.olc-novnooo-o.oo. Vi

l.l.

102.

PART 1. REACTOR DESIGN STUDIES

NUCLEAR CALCULATIONS AND DESIGN STUDIES .seeesecoscosnses

Nuclear CalculationsS eeessseseseersscasssasascsscssacans
Effect of lIon-Exchange Processing of Rare-FEarth
Fission Products on Performance of Interim
Design Reactor Fueled with U233 L..ivevieveennnnns 3
Nuclear Performence of a One-Region, Graphite-
Moderated, Unreflected, Thorium-Conversion,
Molten-Salt-Fueled Reactor seicecececnscacnsennes 3
Initial Nuclear Performance of One-Region,
Graphite-Moderated, Graphite~Reflected, Thorium-

Conversion, Molten-Salt-Fueled Reactors ......... 9
Nuclear Performance of Two-Region, Graphite- |
Moderated, Molten-Salt~Fueled, Breeder Reactor .. 10
Effects of Fast Neutron Reactions in Bed on
Reactivity of Molten-Salt Reactors veeeesececess. 1k
Oracle Code MSPR~Cornpone 020 .teseeseccscossacasese 20
Design of 30-Mw Experimental ReacCtor secsieececccssnsccce 21
F.llel Drain System L I SN BN I IR BN B I BN N Y Y IR N S RN R BN B N R BN OEE BN N R B B N N R R Y 22
Gamma Heating in the Core Vessel of the 30-Mw
E}(perimental Reactor 4 % ® 58 %9 0 F SR TE SRS S S SR 22
COMPONENT DEVELOPMENT AND TESTING «vecveoveevocscasnenes ol
Salt-ILubricated Bearings for Fuel PUMDPS ..coevecrevceeese 2l
Hydrodynamic Journal Bearings seeeceecsessasavoness oL
Hydrodynamic Thrust Bearings ..eceeesvesscscesnsesss 25
Test of Pump Equipped with One Salt-Iubricated
Journal Bearing ® 8 8 8 5 0 PS80 s RS ETEN PSR ea R 25
Bearing Mountings L I B N I BN BN BN DN BN BN BN BN BN OBNE BN IR TN TR BN BN BN NN I N B B B BN BN BN B 25
Mechsnical Seals for Pumps «vvevnees tacestauscacnscserns 27
leEndl}I‘ance Testing LI TN I B N R SR I NN R B I R R N RN I N Y RN R B B N B B RN R N R ) 27
Frozen’-LeadP‘Jmp Seal . 8 B0 B 8 O ¥ & 8N C TSNP R TS S ESeEE ERY e 27
Techniques for Remote Maintenance of the Reactor
Sys-tem e & 2 0 0 & X 0 8D 0 O RS I RS SR AN ER RSN s R eSS ST Ne e s 28
Design, Construction, and Operation of Materials
Testing I.OOPS LR I N I R N I BN B N B NN NN B N B RN O BN BN R BRI BE B RN R RN I N N BN B 32
Forced-Circulation LOODS tevesesrcosrcscscatanocona 32
In-Pile I-oops & 0 & & 8 O 8 B s 9 B S 3G9 P LB SR RO SRS YRR e 35

iii

1‘30 ENGINEERING RESEARCH * 0 5 & % %5 5 0 e 0 ® © @ © 6 % 6 8 8 3 O PG I OO C OSSO0 s o

Physical Property Measurements ...ccevecececccecoscennns
Enthalpy and Heat Capacity ececececsssccvconccscsne
ViSCOSItY seceocnsecsssesossscscnssassssconcsssoncs
Surface Tension +...ee-0a A T

Heat-Transfer StudiesS ceececcscsessescscsocesoncssonnee .
Hydrodynamic StudieS ceeecsessscsosssssccoccsscsscsscacns
1.k, INSTRUMENTS AND CONTROLS eececenssccasscecsssccannsaceos
Molten-Salt-Fuel Level Indicators sceeecccensoscsccscsoas
INOR-3 High-Temperature Pressure Transmitters c¢.c.coeo..

PART 2. MATERIALS STUDIES

D.1. METATLLURGY «vvevececnnencans ettt et e,

Dynamic Corrosion Studies ¢ecceescerscsccrscccscnccccans
INOR“'8 Themal—convection LOOPS @ 9 e 35 8 0008 00400 e
Inconel Thermal—convectiOn LOOPS ® ® 404000 0e0 0o s

General Corrosion StUdiesS ceeeeeccvcrsesencecssacssosnans
Penetration of Graphite by Molten Fluoride

SALTS covsescsnsorssssassscsssesssosncssaccns sessance
Uranium Precipitation from Molten Fluoride Salts

in Contact with Graphite ......... cevscssesescana
Thermal-Convection-ILoop Tests of Brazing Alloys

in Fuel 130 teevecsennenee ceeescescesrecnsenenee
Thermal-Convection-Loop Tests of the Compati-

bility of INOR-8, Graphite, and Fuel 130 ....... .

Mechanical Properties of INOR-8 secveevnronnsocsansannns
Creep TesSTS teeescecetecesonsonsoscsscossnsccscsnans
Fatigue Studies ceeeeeecnacaeocenne cececanns toesnoas
Shrinkage Characteristics of INOR»S ...............

Materials Fabrication Studies ....cc.ccc.. cecsscssasnens
Effect on INOR-8 of Aging at High Temperatures ....
Triplex Heat Exchanger Tubing .cesceces.. ceccacnnna

Welding and Brazing StudiesS eceeeeececcsosnsosscscncecrsos
Procedures for Welding INOR-8 ...vivescevcccocncas .
Mechanical Properties of INOR-8 WeldS .euevevecnesen
Fabrication of Apparatus for Testing the Compati-

bility of Molten Salts and Graphite e.oeeveeceees

2.2, CHEMISTRY AND RADIATION DAMAGE +'vevenerenennnnsncnensns

Phase Equilibrium StudiesS .eeoececsvonsccosoccscsosannsss
The System LiF-BeFQ-ThF4 Secs s et usessssnsessnssnnns
The System NaF-BeFQ-ThFu Ctessecesasssnessensetsence
The SystemNaF-ThFh—UFu ecceccsacsscacace crcenvesaaca

iv

IIl:h.e System SDFE-NHLLIFQ € 5 8 6 6 08 8 00606008800 0C000CePIROSEIOEILE
Solubility of PuF3 in Converter Fuels ......... covas
Separation of LifF from LlTF—Bng ............ cosses
Fission-Product Behavior ...... cessceen s sesasacaes cesesee
PI‘ECipitation Of SmF3 With CEF3 8 @ 9 06 006 00 000 s s 00 ¢ e s 0
Chemical Reactions of Oxides with Fluorides in
Molten-Fluoride-Salt Solvents ..cceeceveccecans oo
Gas Solubilities in Molten Fluoride Salts ...cvvececccess
Solubility of Neon in LiF--BeF2 ....... crsesscssenncs
Solubility of COp in NaF-BeZF2 ceciscansecsssssccsse e
Chemistry of the Corrosion Process ......ceoee eeasee ceees
Samples from Operating Loops ..ceeceeciiereeenaanss .o
Radioactive Tracer Analyses for Iron in Molten
Fluoride Salts covvveeecceecnnas cesecscensscncsans .
Activities in Metal AllOysS .ececcecens cscacsescnass .o
- Vapor Pressures of Molten Salts ........ Ceesseassesessens
Permeability of Graphite by Molten Fluoride Salts .......
Radiation Damage Studies «.ccievceen crsecsacsnrancens sreane
INOR-8 Thermal Convection Loop for Operatlon in
the LI'PR- .......................... & 8 9 & & 5O s 00 0t
In-Pile Static Corrosion Tests .eeeeeieccens D
Preparation of Purified Materials ............ cesceas ceee
Purification, Transfer, and Service Operations teces
Fuel Replenishment Tests +iceeen... cesccae
Pure Compounds Prepared with Molten Ammonlum
Blfluorlde ® @ ® 6 8 4 6 % 6 6 S B YV S e e O e s a0 o ¢ & & " & 0 9 6 8 0
Reaction of Chromous Fluoride with Stannous
Fluoride .............. ¢ 6 & ® & 2 O & s s 0 e 0 0 & & 8 & % & O & & 0
FUEL PROCESSING ...... ceeieas Ceeeneaas

2.3,

|

MOLTEN-SALT REACTOR PROGRAM QUARTERLY PROGRESS REPORT
SUMMARY

Part 1. ZReactor Design Studies

1.1. Nuclear Calculations and Design Studies

The effects of fuel processing by the ion-exchange method at a
rate of once per year and at a rate of 12 times per year were compared
for the Interim Design Reactor fueled with U233. Processing at the
higher rate gave insufficient improvement in performence to offer a |
substantial economic advantage.

The nuclear performasnce of a one-region, graphite-moderated,
unreflected, thorium—conversibn, molten-salt-fueled reactor was
studied to evaluate the effects of processing the fuel at the rate
of once per year and 12 times per year, the effects of decreasing
the thickness of the core vessel and adding a blanket, and the effects
of using a graphite core vessel and a blanket. The results of the
calculations indicated that the nuclear performence of the one-region
reactor fueled with Ug35 and processed by either the volatility or
ion-exchange methods at the rate of one fuel volume per year is com-
petitive with the best performance of the solid fuel systems, and
therefore the molten-salt reactor will yield lower over-all fuel costs
when the potential economies of continuous, chemical processing of
the fluid fuel are realized.

The initial nuclear characteristics of a one-region, heterogeneous,
- graphite-moderated and -reflected, thorium-conversion, molten-salt-
fueled reactor were studied. It was found that with 13 mole % ThF) in
the fuel, a maximum regeneration ratio of 0.846 could be obtained
with an inventory of approximately 1150 kg of U5°. With 7 mole %
ThFu in the fuel, an estimated maximum regeneration ratio of 0.798 at

U235 was obtained.

an inventory of approximately 900 kg of
Calculations were also made of the performance of two-region,

graphite-moderated, molten-salt-fueled, breeder reactors. OSpherical

vii

reactors, 5 ft in diameter, with fuel volume fractions of 0.10 and
0.15% were studied. Results were obtained for reactors with power
levels of 125 and 250 Mw. It was found that the only reactor with
a doubling time less than 20 years was the 250-Mw reactor having a
fuel volume fraction of 0.10. 5

on the reactivity

9

of molten-salt-fueled reactors were studied. It was found that Be

The effects of fast-neutron reactions in Be

is an appreciable poison in homogeneous, molten-salt-fueled reactors,
but it has negligible effect in graphite-moderated, molten-salt-

fueled reactors.
Work is continuing on the design of a 30-Mw, one-region, experi-

mental, molten-salt reactor.

1.2. Component Development and Testing

Tests of salt~-lubricated hydrodynamic journal bearings were
continued in order to obtain operating data on which to base an
optimum design. Assembly of a thrust-bearing tester was nearly com-
pleted. An existing centrifugal sump pump is being modified for
service tests of a salt-lubricated journal bearing, and means for
flexibly mounting the bearing to the pump casing are being investigated.
The modified Fulton-Sylphon bellows-mounted seal being subjected
to an endurance test in a PK-P type of centrifugal pump has continued
to seal satisfactorily forvmore than 12,000 hr of operation. Operation
of an MF type of centrifugal pump with fuel 30 as the circulated
fluid has continued to be satisfactory through more than 15,500 hr,
with the past 13,500 hr of operation being under cavitation damage
conditions. A small frozen-lead pump seal on a 3/16-in.-dia shaft
has operated since the first 100 hr of 7500 hr of operation with no lead
leakage. A similar seal on a 3 1/b-in.-dia shaft has operated 3600
hr, with an average leakage rate of 9O cms/hr. Operational data suggest
that the seal should be redesigned to provide better coolant control
and packing to decrease the annulus between the seal and the shaft.
Construction work continued on the remote maintenance demon-

stration facility. The work is on schedule and is to be completed by

viii

June 30, 1959.

The operation of long-term forced-circulation corrosion-testing
loops continued in 15 test stands. Tests of two Inconel loops that
had operated for one year were terminated. GSalt samples were removed
periodically from two loops that contain sampling devices.

The in-pile loop operated previously in the MIR was disassembled.
It was found that oil which had leaked past the pump shaft seal and
filled the oil trap on the purge outlet line was polymerized by radia-
tion, and the polymerized oil plugged the outlet line. The second
in-pile loop, which was modified to minimize the probability of purged
line plugging and activity release, was installed in the MIR on
April 27.

1.3. Engineering Research

The enthalpy, heat capacity, viscosity, and surface tension have
been experimentally obtained for several additional fluoride salt

mixtures containing BeF, with varying amounts of UF) and/or ThF), .

Altering the salt compoiition toward higher percentages of the high-
molecular-weight components resulted in substantial decreases in both
the enthalpy and heat capacity. The viscosity showed some increase

(of the order of 10%) as the percentage of UF), and ThF) in the mixtures
was increased. Initial flow calibrations of a full-scale mockup of

the pump system designed for the study of interfacial film formation

and heat transfer with BeF,.,~-containing salts have been completed;

2
assembly of the components is proceeding. The flow characteristics
of the sintered-metal-filled annulus of a double-walled tube have been

obtained for two porosity conditions.

1.k. Instruments and Controls

Work has continued on molten-salt-fuel level indicators. Two
Inconel "I'"-tube-type elements were prepared for testing, and an
INOR-8 element and test vessel are being constructed.

Six INOR-8 pressure transmitters and indicating systems were

ordered for testing. These units are of the pneumatic~-indicator type.

ix

Part 2. Materials Studies

2.1. Metallurgy

Corrosion studies were completed on four INOR-3 and four Inconel

thermal-convection loops. Metallographic examination of the INOR-3
loops, three of which operated for 1000 hr and the other for one year,
revealed no observable attack. The Inconel loops, which represented
three one-year tests and one 1000-hr test, showed intergranular void
attack to depths ranging from L to 15 mils. Twelve thermal-convection-
loop tests were initiated during this period. The scheduled tests of
two forced-circulation loops were completed, and one new forced-
circulation-loop test was started.

The effects of penetration of graphite by molten~fluoride-salt
fuels are being investigated in order to evaluate the problems
associated with the use of unclad graphite as a moderator. In static
pressure penetration tests at 150 psia and lSOOOF, partially degassed
graphite was penetrated throughout by fuel 30 (NaF-ZrFu—UFh) but was
not penetrated by fuel 130 (LiF-Bng-UFu), as indicated by macroscopic
examination. Thermal cycling of the graphite that was penetrated
by fuel 30 did not result in damage to the graphite. Additional static
and dynamic tests are planned.

A series of tests was run to further investigate the precipitation
of uranium from fuel 130 held in a graphite crucible at lBOOOF. The
results of these tests supported the previous conclusion that the
uranium precipitation was the result of the fuel reacting with oxygen
supplied by degassing of the graphite.

Several brazing alloys were tested for compatibility with fuel
130. Coast Metals alloy No. 52 (89% Ni-4% B-5% Si~2% Fe) and pure
copper brazed to Inconel and INOR-8 showed good resistance to fuel
130 in 1000-hr tests at 13OOOF in a thermal-convection loop. The
following brazing alloys showed some attack or porosity as a result
of similar exposure to fuel 130:‘ General Electric alloy No. 81
(70% Ni—19% Cr—11% Si), Coast Metals alloy No. 53 (81% Ni-8% Cri% Si— -
L% B-3% Fe), and a gold-nickel alloy (82% Au—18% Ni).

A specimen of INOR-8 was examined for evidence of carburization
that had been exposed to fuel 130 for 4000 hr in a thermal-convection
loop hot leg at a temperature of l3OOOF. The presence of a 10-in.
graphite insert in the hot leg of this loop did not cause carburization
of the INOR-8.

No new creep data for INOR-8 became available during the quarter.
Creep tests are presently in progress at 1100 and 1200°F with the
fluoride salts of interest as the test environments, and these tests

are expected to run in excess of 10,000 hr.
The critical results obtalined from rotating-beam fatigue tests

at 1500°F indicate that INOR-8 has significantly better fatigue
resistance than Inconel. Creep and reiaxation tests indicate that an
unstable condition exists in the temperature range from 1100 to 1400°F.
There is some indication that a second phase appears which causes
contraction of the metal even under load.

Based on the results of tensile tests conducted on specimens of
INOR-8 aged for 10,000 hr in the temperature range of 1000 to 1400°F,
it has been concluded that INOR-8 does not exhibit embrittling tendencies
that can be attributed to high-temperature instability. No significant
differences were found between the tensile properties of the aged
specimens and those of specimens in the annealed condition.

An effort is now being made to fabricate triplex heat exchanger
tubing containing a prefabricated porous nickel core. Porous nickel
tubes have been ordered from Micro Metallic Corp. to determine the feasi-
bility of cladding the material with Inconel and INOR-8. A sample piece
of porous nickel has been incorporated into the annular space formed
between two Inconel tubes for conducting a preliminary cladding experiment.

Procedures are being developed for fabricating INOR-8 material
ranging in size from thin-walled tubing to heavy plate. The procedures
thus developed are being qualified in accordance with methods prescribed
by the ASME Boiler Code.

The effects of various deoxidation and purification processes

on the mechanical properties of INOR-8 weld metal are being studied

xi

in an effort to improve the high-temperature ductility of INOR-8 weld
metal. Several heats of weld metal containing various additives were
cast and fabricated into weld wire for mechanical property evaluation.
A method was developed for brazing graphite to Inconel. A
commercially available brazing alloy composed of silver, titanium, and
copper was found that wet vacuum-degassed graphite. Such graphite-to-
Inconel joints were used in the fabrication of equipment for studying
the penetration of graphite by molten salts in a dynamic, high-pressure

system.

2.2. Chemistry and Radiation Damage

A revised phase diagram for the system LiF-BeFE-ThFu was prepared
that includes new data obtained from thermal-gradient quenching experi-
ments. Quenched samples from experiments in which an equilibration
period of 3 weeks was used revealed that the area of single-phase
ternary solid solutions involving 3LiF~ThFu is greater than previously
reported. A phase diagram showing the progress made thus far in the
study of the NaF-Bng-ThFu was also prepared, and the identity and
approximate locations of primary phases in the system N&F~ThFu-UFA
were determined.
| The SnFe--NHuHF2 system was investigated because of its potenti-
alities as a strongly oxidizing, low-melting solvent for reprocessing
fuels. Reliable data were obtained only in the range O to 40% San.
A 15 mole % addition of SnF, gives a mixture with a melting point of -
about 100°C.

Measurements were made of the solubility of PaF? in LiFnBng-

2

UFh (70-10-20 mole %). The solubility values obtained were all higher

than those obtained with LiF-BeF, mixtures having about the same LiF

concentration but no UF&' °

The possibility of separating LiF from the mixture LiF-BeF,, (63~
37 mole %) by adding NaF and decreasing the temperature was investigated.
In an initial experiment only 23% of the LiF remained in solution when
the temperature of the mixture to which NaF was added at TOOOC was

lowered to L490°C.

xii

Tests were initiated for determining the rate of exchange in a
proposed method for decreasing the total rare earth content of molten

fluoride salts. The exchange reaction

SmF3 (a) + CeF3 (s)“‘fiCer (d) + Sn@3 (s)

is utilized by passing the salt through an isothermal bed of solid
CeF3 to lower the SmF3 content, and then the temperature of the effluent

salt is lowered to decrease the total rate earth content.

Two methods for separating uranium from fission products are

- being studied. In one method the reaction of UFh with BeO to produce

UO2 is utilized. In the other method, the reaction of UFA with water

vapor produces UO The sharpness of the separation with water vapor

may be advantageois in processing schemes. It is thought that a
process can be developed that will eliminate the need for fluorine
and that will provide for the simultaneous removal of uranium and
thorium. A further step would be required to remove the rare earths.

Measurements were made of the solubility of neon in LiF-BeF
and CO, in NaF-BeF,..

2 2
In the study of the chemistry of the corrosion process, further

2

samples of melts from operating INOR-8 and Inconel forced-circulation
loops were analyzed for chromium. The chromium concentration in the
LiF-BeFE-THthUFu mixture in the INOR-3 loop reached a plateau of
about 550 ppm after about 1200 hr of operation. The chromium concen-
ration of the same salt mixture in the Inconel loop increased more
rapidly than in the INOR-8 loop and after 1700 hr was still increasing.
These tests are continuing.

Most investigations of corrosion behavior depend on accurate
analyses for structural metel ions, and therefore anomalies in the
present analytical methods are being studied.

Tests of the permeability of graphite by molten fluoride salts
have continued. Three types of graphite that were specially treated
to meke them impervious were obtained from the National Carbon Company.

The types designated ATL-82 and ATJ-82 were resistant to forced

xiii

impregnation with LiF-MgFg salt and were considerably resistant to
penetration by a typical reactor fuel. Unexpectedly high concentrations
of uranium in the center of the rods are being investigated.

The in-pile thermal-convection loop for testing fused-salt fuel
in INOR-8 tubing in the LITR was operated in preliminary fests out-of-
pile, and satisfactory circulation was obtained. Final assembly of the
loop system is under way. Two fuel-filled INOR-8 capsules were installed
in the MTR and are being irradiated at lQBOOF.

A device for testing a proposed fuel sampling and enriching
mechenism was constructed and tested. It was found that the rate of

solution of solid UFM in LiF-BeF, would be adequate for convenient

enrichment procedures. °
The use of molten ammonium bifluoride as a reactant for preparing

both simple and complex fluorides and the preparation of pure chromous

fluoride by the reaction of chromium metal with molten stannous

fluoride were studied,

2.3. Fuel Processing

Studies of the processing of molten fluoride salt fuels by the
fluoride~-volatility process were continued. Further measurements of
the solubility of neptunium (1IV) in aqueous solutions saturated with
LiF-BeF ““HFM'UFM indicate that a solubility of the order of 0.0002

2
to 0.00005 mole % may be expected in actual processing.

Xiv

PART 1.

REACTOR DESIGN STUDIES
1.1. NUCLEAR CALCULATIONS AND DESIGN STUDIES

Nuclear Calculations

Effect of Ion-Exchange Processing of Rare-Earth Fission Products on
233

Performance of Interim Design Reactor Fueled with U

It was reported previouslyl that substantial savings in fuel
burnup and inventory couid be achieved in the Interim Design Reactor2
fueled with U235 by passing the fuel salt rapidly through beds of
CeF3 to remove the rare earth fission products. The effect of ion
exchange processing on the performance of the same reactor system
fueled with U233 has now been studied. The results for two different
processing rates — once per year and 12 times per year — are compared
in Fig. 1.1.1. Processing at the rate of once per year (1.7 ft3/day)
could bé performed either by the fluoride-~volatility methodl or by
the ion-exchange method. Because of the high cost associated with the
discarding of carrier salt in the volatility method, it is not feasible
to use the volatility process for the higher rate.

It may be seen that processing at the higher rate gives only a
small improvement in performance. It is doubtful therefore that the
ion-exchange process offers any substantial economic advantage in the

233

Interim Design Reactor system fueled with U

Nuclear Performance of a One-Region, Graphite-Moderated, Unreflected,
Thorium-Conversion, Molten-Salt-Fueled Reactor

The nuclear performance without processing to remove fission
products of a reactor having a spherical core 1 ft in diameter and
fuel channels 3.6 in. ID arranged in a square lattice on 8 in. centers,
was described previously.3 The performance of the same system with

various processing rates and with other modifications has now been

MSR Quer. Prog. Rep. Jan. 31, 1950, ORNL-268k, p 25.

®Molten Salt Reactor Program Status Report, ORNL-2634% (Nov. 12,
1958).

3 2681

MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL- , p 31.

UNCLASSIFIED
ORNL-LR-DWG 38445

1.08 |
o \ PROCESSING RATE, FUEL VOLUMES/year
5 1.06 a—— —
0’4 \ 12
=
o T — 4
5 1.04
e
uJ
uZ_l ‘
o 1.02
!
o
1.00
2000
INVENTORY A
“w 1500 S—
w
E A
0
o]
5
.§ 1000 CORE DIAMETER: 8 ft
s REACTOR POWER: 600 Mw {th), 250 Mw (e)
= PLANT FACTOR: 0.8
> VOLUME OF FUEL: 600 ft°
O FUEL: U%®®
5 500 UEL
bl
>
=
.
<
&)
=
% CUMULATIVE NET BURNUP
0 . m—
——— 1
I
12
-500
0 4 8 12 16 20

TIME OF OPERATION (yr)

Fig. 1.1.1. Effect of Fuel Processing Rate on Nuclear Performance of Spherical, Homogeneous, Two-
Region, Molten-Salt-Fueled, Breeder Reactor.

studied, and typical results are presented in Table 1.1.1 and in
Fig. 1.1.2, where regeneration ratio; eritical inventory, and net
cumulative burnup are plotted as functions of time of operation.
Without processing, the regeneration ratio falls from 0.8 to 0.5 in
20 years. The fuel additions required to override fission-product
poisons are large; the critical inventory rises from 800 to 1800 kg
of fuel, giving an average specific power of only 0.4 Mw/kg. The
burnup is also large, averaging 0.45 g/Mwd.

Processing the fuel salt to remove fission products at the rate
of one fuel volume per year (2.5 ftB/day) stabilizes the regeneration
ratio at above 0.8. The inventory averages about 750 kg. The burn-
up 1s reduced to 0.21 g/de. Increasing the processing rate to 12
fuel volumes per year (30 ft3/day) improves the performance only
slightly.

Although it is uncertain that a core vessel and blanket could
be incorporated into this reactor, it was of interest to evaluate
the nuclear benefits of such an arrangement. Accordingly, the thick-
ness of the INOR-8 reactor vessel wall was reduced to 1/2 in. for
the calculations, and a 30-in. blanket containing 13 mole % ThF) in
a mixture of LiF and BeiF2 was added. The blanket system was processed
12 times a year by the fluoride-volatility method to remove U233,
which was added to the fuel salt. The fuel was processed 12 times
a year by the ion-exchange method to remove fission products. The
results of the calculations are presented in Table 1.1.2, and the
performance is indicated in Fig. 1.1.2 by the lines labeled U235 12,B.
The regeneration ratio averaged in excess of 0.96 and approached 1.0
in the steady state. The critical inventory was not decreased
significantly, but was shifted toward u?33 in composition (90%). The
burnup was reduced to 0.056 g/MWd, and the reactor was practically
self-sustaining. However, it was estimated that the savings in mills
per kilowatt-hour effected by reducing the burnup costs would be more
than offset by the additional capital costs of reactor vessel, blanket

materials, pumps, processing equipment, and concomitant operating costs.

Table 1.1.1. Effect of Fuel Processing Rate on Nuclear Performance of One-Region, Graphite-Moderated,

Unreflected, Thorium-Conversion, Molten-Salt-Fueled Reactor

Core diameter: 14 ft

Reactor power: 600 Mw(th), 250 Mw(e)
Plant factor: 0.8

Fuel volume: 900 3

Fuel channels: 3.6 in. ID

Lattice: triangular, 8-in. centers
Core vessel: l-in.-thick INOR-8

Blanket: none

o s After 20 Years with Fuel After 20 Years with Fuel
Processed Once per Year  Processed 12 Times per Year
Inventory Absorption
(kq) Ratio® Inventory Absor?f:]on Inventory Absor.ptti]on
(kg) Ratio (kg) Ratio
Fissionable isotopes
y233 563 0.764 559 0.794
4235 829 1.000 187 0.219 167 0.205
pu2¥ 3 0.017 0.3 0.001
Fertile isotopes
Th232 38,438 0.783 38,438 0.776 38,438 0.807
u234 218 0.051 222 0.054
4238 64 0.008 142 0.017 136 0.017
Fuel carrier
Li’ « 6,328 0.054 6,328 0.052 6,328 0.055
19 37,571 0.025 37,571 0.025 37,571 10.026
Moderator
Be’ 1,840 0.001 1,840 0.001 1,840 0.001
cl? 0.051 0.049 0.052
Fission products 181 0.034 15.3 0.003
Parasitic isotopes
U3 and others 197 0.037 184 0.036
Miscellaneous
po233 17 0.009 17.8 0.010
Core vessel and 0.149 0.151 0.155
leakage
Neutron yield, 1 2.071 2.207 2.217
Total fuel inventory, kg 829 754 726
Cumulative net burnup, kg 0 757 686
Net fuel requirement, kg 829 1,51 1,412
Regeneration ratio 0.791 0.836 | 0.8683

“Neutrons absorbed pet neutron absorbed by fissionable isotopes.

UNCLASSIFIED
ORNL—LR—DWG 38146

(12)

\
W
/

o
®

7
==

T~

REGENERATION RATIO

\
CORE DIAMETER: 14 ft ~—

REACTOR POWER 600 Mw (th), 250 Mw (e) ——
PLANT FACTOR: 0.8

0.4 | FUEL VOLUME: 900 ft3
FUEL CHANNELS: 3.6in. ID
2000  LATTICE: TRIANGULAR, 8-in. CENTERS 1 I |

/

FUEL AND PROCESSING RATE
(fuel volumes /year)

U235 (O)V
1500 — .

1000 » -

/"
A
\

S ———
e ————
P ——

500

CUMULATIVE NET

BURNUP /
,f/
O P ——c——

N
\

/

Efi

CRITICAL INVENTORY (kg of fissionable isotopes )
c\
[\¥}
(&}
Pt
I
\‘\[E '

C
N
(&)
w

o

w

—500 1
8 10 12 14 16 18 20

TIME OF OPERATION (yr)

O
N
N
o

Fig. 1.1.2. Nuclear Performance of Spherical, Heterogeneous, Graphite-Moderated, One-Region, Unre-
flected, Thorium-Conversion, Molten-Salt-Fueled Reactors.

Table 1.1.2. Effect of Core Vessel Material and Blanket on Nuclear Performance of One-Region,
Graphite-Moderated, ThoriumeConversion, MoltenSalt-Fueled Reactor -

Core diameter: 14 ft

Reactor power: 600 Mw(th), 250 Mw{e)

Plant factor: 0.8 -
Fuel volume: 900 t3

Fuel channels: 3.6 in. ID

Lattice: triangular, 8-in. centers

Blanket thickness: 30 in.

Fuel processing rate: 12 times per year

Reactor with 1/2-in.-Thick INOR-8 Reactor with Graphite
Core Vessel Core Vessel
Initial State After 20 Years Initial State After 20 Years
Inventory  Absorption Inventory Absorption Inventory  Absorption Inventory Absorption
(kg) Ratio® (kg) Ratio® (kg) Ratio® (kg) Ratio®
Fissionable isotopes
U233 (fuel) 625 0.917 594 1.000 620 0.936
U233 (blanket) 1.7 3
u23s 829 1.000 65 0.082 48 0.064
Py23 0.2 0.001
Fertile isotopes
Th232 (fyel) 38,438 0.783 38,438 0.831 38,438 0.933 38,438 0.852
Th232 (blanket) 76,765 0.095 76,765 0.101 76,765 0.156 76,765 0.149
u234 239 0.060 277 0.072
u2s3e 64.3  0.008 104 0.013 1
Fuel carrier
Li’ 6,328 0.054 6,328 0.057 6,328 0.068 6,328 0.060
F19 37,571 0.025 37,571 0.026 37,571 0.028 37,571 0.027
Moderator
Be’ 1,840 0.001 1,840 0.001 1,840 0.001 1,840 0.001
c'? 0.051 0.054 0.063 0.056 .
Fission products 15 0.003 15.3 0.003
Parasitic isotopes
U236 and others 122 0.026 20 0.003 .
Miscellaneous
Pa233 (fuel) 18.1 0.010 18.6 0.011
Pa233 (blanket) 2.2 3
Blanket carrier salt, core 0.055 0.058 0.006 0.061
vessel and leakage
Neutron yield, 7 2.071 2.240 2.256 2.241
Total fuel inventory, kg 829 691 594 670
Cumulative net burnup, kg 0 185 0 245
Net fuel requirement, kg 829 876 594 425
Regeneration ratio 0.885 0.995 1.090 1.064

°Neutrons absorbed per neutron absorbed by fissionable isotopes.

In a further extension of the calculations, the nuclear benefits
to be obtained by the use of a graphite core vessel (in place of INOR-8)
233
U

and fueling with were studied. These changes are, of course,
impractical in the present state of technology. However, the study
was performed in order to define the limiting nuclear performance of
this particular core and fuel system.

The results of the calculations are presented in Table 1.1.2, and
the performance is indicated in Fig. 1.1.2 by the lines labeled U233
12,B. The regeneration ratioc is stabilized above 1.06, the average
inventory is about 650 kg, and the cumulative net burnup is slightly
negative, amounting to -250 kg of U233 in 20 years. However, 76 kg
of this is required as increased inventory in the fuel salt to override
fission products and nonfissionable isotopes.

It is concluded that the nuclear performance of the one-region
reactor fueled with U235 and processed by either the volatility or
ion~-exchange methods at the rate of one fuel volume per year is
competitive with the best performance of the solid fuel systems, and
therefore the molten-salt reactor will yield lower over-all fuel costs
when the potential economies of continuous, chemical processing of
the fluid fuel are realized.

Initial Nuclear Performance of One-Region, Graphite-Moderated, Graphite-
Reflected, Thorium-Conversion, Molten-Salt-Fueled Reactors

The initial nuclear characteristics of a heterogeneous, graphite-
moderated and -reflected, one-region, molten-salt-fueled reactor were
studied. The reactor considered in the study consists of a cylindrical
core, 15 ft in dia and 15 ft high, surrounded by a 2.5-ft-thick graphite
reflector contained in an INOR-8 pressure shell 1.5 in. thick. The
core is penetrated by cylindrical fuel passages arranged in an 8-in.
triangular lattice parallel to the core axis. The resulting unit cells
are hexagonal and 15 ft long. This system has been investigated over
a range of fuel volumetric fractions in the core and at two concen-
trations of thorium in the fuel salt. The initial nuclear character-

3

istics of the system, with an externsal fuel volume of 673 ft-, are

given in Tables 1.1.3 and 1.1.4 for the cases having 13 and 7 mole %
ThFu in the fuel, respectively; Fig. 1.1.3 shows the regeneration ratiq g
as a function of the system inventory.

The modified Oracle program Cornpone was used to calculate the
multiplication constant and group disadvantage factors for the fuel
and graphite. The results were then used for complete reactor calcu- |
lations in spherical geometry on a core having the same volume as the
cylindrical core. Mean, homogeneized densities (atoms/cm3) were used
for each element.

It may be seen from Fig. 1.1.3 that, in the case having 13 mole %
ThFu in the fuel, a meximum regeneration ratio of 0.846 is obtained
at an inventory of approximately 1150 kg of U235. With 7 mole % ThFh
in the fuel, an estimated maximum regeneration ratio of 0.798 at an
inventory of approximately 900 kg of U235 was obtained.

Nuclear Performance of Two-Region, Graphite-Moderated, Molten-Salt-
Fueled, Breeder Reactor

The nuclear performance of spherical, graphite-moderated reactors,
5 ft in diameter with fuel volume fractions of 0.10 and 0.15, was
studied. The fuel channels were 2.6 in. in diameter and were arranged
in an 8-in. trianguler lattice. These reactors were surrounded by
2-in.-thick graphite core vessels and 30-in.-thick blankets.

The fuel consisted of 71 mole % LiF, 16 mole % BeF,, and 13 mole %
ThFu plus UFM‘ The blanket salt had the same composition as the fuel,
but no UFM'

The performance of these reactors over a period of 20 years was
calculated for power levels of 125 and 250 Mw. The processing rate
in all cases was 12 fuel volumes per year. The system volumes used
for these calculations are shown in Table 1.1.5; only the heat exchanger
volume was increased for higher power levels.

The results of the calculations are presented in Tables 1.1.6 and
1.1.7 and are plotted in Figs. 1.1.4, 1.1.5, and 1.1.6. It may be seen
that the only reactor with a doubling time less than 20 years is the o

250-Mw reactor having a fuel volume fraction of 0.10. For both cores,

10

1T

Table 1.1.3. Initial Nuclear Performance of a One-Region, Graphite-Moderated, Thorium-Conversion, Molten-Salt-Fueled Reactors
with 13 mole % ThF4 in the Fuel

Fuel Volumetric Fraction in the Core

0.12 0.16 0.20 0.22 0.24 0.28
Inventory Absorption |nventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption
(kg) Ratio? (kg) Ratio (kg) Ratio (kg) Ratio tkg) Ratio (kg) Ratio
Fissionable isotope
u 233 798.3 1 925.3 1 1,086 ! 1,168 1 1,299 1 1,526 1
Fertile isotopes
Th232 42,338 0.804 46,883 0.820 51,408 0.837 53,671 0.836 55,954 0.831 60,480 0.823
u238 59.9  0.005 69.4  0.006 81.4  0.008 87.6  0.009 97.4  0.011 114.5  0.012
Fuel carrier
Li’ 6,974 0.064 7,723 0.061 8,469 0.055 - 8,841 0.055 9,217 0.049 9,963 0.046
19 41,350 0.023 45,789 0.026 50,208 0.026 52,418 0.028 54.648 0.027 59,068 0.030
Moderator
Be? 2,021 0.001 2,238 - 0.001 2,454 0.001 2,562 0.001 2,674 0.001 2,887 0.001
cl? 334,900 0.079 328,900 0.056 323,000 0.040 320,000 0.036 317,000 0.030 311,000 0.025
Core vessel capture 0.105 0.104 0.089 0.089 0.093 0.091
plus leakage
Neutron yield, 7 2.081 2.074 2.056 2.054 ©2.042 2.028
Regeneration ratio 0.809 0.826 0.845 0.845 0.842 0.835

9Neutrons absorbed per neutron absorbed in fissionable isotope.

A

Table 1.1.4. Initial Nuclear Performance of a One-Region, Graphite-Moderated, Thorium-Conversion, Molten-Salt-Fueled Reactors

with 7 mole % ThF4 in th

e Fuel

Fuel Volumetric Fraction in the Core

0.12 0.16 0.20 0.22 0.24
Inventory  Absorption Inventory  Absorption Inventory  Absorption Inventory  Absorption Inventory  Absorption
(kg) Ratio® (kg) Ratio (kg) Ratio (kg) Ratio (kg) Ratio
Fissionable isotope
u 235 472 ] 506 592 1 633 1 677
Fertile isotope
Th232 22,926 0.710 25,387 27,837 0.769 29,060 0.771 30,300
u238 35.3 0.003 37.3 44.8 0.005 46.9 0.006 50.3
Fuel carrier
Li7 6,641 0.105 7,355 8,065 0.097 8,420 0.099 8,777
F19 39,440 0.028 43,670 47,890 0.032 50,000 0.034 52,120
Moderator
Be? 3,284 0.001 3,637 3,988 0.001 4,163 0.001 4,349
cl? 334,900 0.135. 328,900 323,000 0.072 320,000 0.066 317,000
Core vessel capture 0.123 0.100 0.101
plus leakage ‘
Neutron yield, 1 2.105 2.076 2.078
Regeneration ratio 0.713 0.760 0.774 0.777 - 0.787

9Neutrons absorbed per neutron absorbed in the fuel.

UNCLASSIFIED
ORNL—LR—DWG 38147

0.86
0.84 // \\\
13 mole % Thf, IN FUEL \o\
0.82
0.80 =
/—/’-— —
o
2 i
o
= /
s}
£ 078
= /o
wJ
=2
1]
e 7 mole 7 ThF IN FUEL
o
0.76 9
CYLINDRICAL CORE: 15 ft DIA, 45 ft HIGH
REFLECTOR: 2.5 - ft~THICK GRAPHITE
PRESSURE SHELL: 1.5~in.~THICK INOR-8
0.74 EXTERNAL FUEL VOLUME: 673 ft°
072
0.70
0 200 400 600 800 1000 1200 1400 1600

U235 INVENTORY (kg)

Fig. 1.1.3. Initial Nuclear Performance of One-Region, Graphite-Moderated, Thorium-Conversion,
Molten-Salt-Fueled Reactors. :

13

Table 1.1.5. Characteristics of Two-Region, Graphite-Moderated,
Molten-Salt-Fueled, Breeder Reactors

Fuel volume fraction

in core 0.10 0.15
Power, Mw 125 250 125 - 250
Fuel volume, ft3

Core 6.67 6.67 10 10
Heat exchanger 15 30 15 30
Other 15 15 15 15

Total 36.7 51.7 Lo 55

Average inventory, kg
of fissionable

isotopes 73 110 76 135
Average regenerstion

ratio 1.090 1.075 1.073 1.068
Annual production,? kg 2.15 4 b5 | 1.7 3.2

a . ' 233
Annual production = annual average excess of U 33 based on a reactor
lifetime of 20 years.

the regeneration ratio drops off at the higher power level. This was

to be expected because of the buildup of U234 and U236

; also, since
the processing rate was the same for all power levels, the concentration
of the fission fragments was increased at the higher power level.
Effects of Fast Neutron Reactions in Be9 on Reactivity of Molten-Salt
Reactors | |

4

The net effect of the reactions Be9(n,2n)2He

, and Beg(n,a)He6(B)Li6

on the reactivity of molten-salt reactors was studied. In a spherical,
two-region, homogeneous core 8 ft in diameter having a carrier salt

composed of 58 mole % LiF, 35 mole % BeF,., and 7 mole % ThF), , and ye33

2 9

fuel, the contribution of the reactions in Be” to the neutron multi-
plication was estimated to be -0.087. In a spherical, two-region,
heterogeneous, graphite-moderated, molten-salt reactor having a core
5 £t in diameter, a carrier salt composed of 71 mole % LiF, 16 mole %

BeF,, and 13 mole % ThF) , and fueled with U233, the reactivity effect

1k

Table 1.1.6, Nuclear.Performance of Two-Region, Graphite-Moderated, Molten-Salt.Fueled,
Breeder Reactors with Fuel Yolume Fraction of 0.10

Core diameter: 5 ft

Lattice: triangular, 8-in. center

Fuel processing rate: 12 times per year

For 125-Mw(th) Plant

For 250-Mw(th) Plant

After 20 Years

Initial State Initial State After 20 Years
Inventory  Absorption  Inventory  Absorption Inventory Absorption  Inventory  Absorption
{kg) Ratio® (kg) Ratio?® (kg) Ratio® (kq) Ratio®
Fissionable isotopes
U233 (fuel) 59.1  1.000 67.9 0.910 83 1.000 99 0.898
U232 (blanket) 2.37 5
u23s 7.8 0.088 13 - 0.098
Py23 0.038 0.002 0.10 0.004
Fertile isotopes : ‘
Th?%2 (fuel) 1,567 0.402 1,567 0.335 2,207 0.402 2,207 0.322
Th2%2 (blanker) 21,527 0.720 21,527 0.660 21,520 0720 21,520 0.648
Y234 39.8 0.092 64 0.100
U238 1.34 0.002 4 0.005
Fuel carrier
Li’ 258 0.027 258 0.020 363 0.027 363 0.019
F19 . 1,532 0.013 1,532 0.012 2,158 0.013 2,158 0.012
Moderator
Be’ 75 0.001 75 0.001 106 0.001 106 0.001
c? 0.042 0.032 0.042 0.031
Fission products 3.18 0.006 6 0.008
Parasitic isotopes
U236 and others 6.21 0.014 13 0.021
Miscellaneous
Pa233 (fyel) 1.51 0.008 3 0.011
Pa233 (blanket) 3.05 6
Blanket carrier salt, core 0.060 0.055 0.060 0.054 -
vessel, and leakage
Neutron yield, 7 2.265 2.238 2.265 2.232
Total fuel inventory, kg 59.1 78.1 83.0 117
Cumulative net burnup, kg 0 ~72.2 0 -123
Net fuel requirement, kg 59.1 5.9 83.0 ~6
Regeneration ratio 1.122 1.122 1.064

1.080

“Neutrons absorbed per neutron absorbed by fissionable isotopes.

15

Table 1.1.7. Nuclear Performance of Two-Region, Graphite-Moderated, Molten-SaltFueled,
Breeder Reactors with Fuel Yolume Fraction of 0.15
Core diameter: 5 ft
Lattice: triangular, 8-in. center
Fuel processing rate: 12 times per year

For 125-Mw(th) Plant " For 250-Mw(th) Plant
Initial State ' After 20 Years Initial State " After 20 Years
Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption
(kg) Ratio® (kg) Ratio® (kg) Ratio® (kg) Ratio®

Fissionable isotopes

U233 (fuel) 59 1.000 72 0.907 103 1.000 127 0.901

U233 (blanket) 2 4

u23s 9 0.090 17 0.095

Py2% 0.1 0.003 ‘ 0.2 0.004
Fertile isotopes

Th232 (fuel) 1,710 0.472 1,710 0.384 2,987 0.472 2,987 0.378

Th232 (blanket) 23,487 0.642 23,487 0.589 23,527 0.642 23,527 0.585

u234 44 0.095 81 0.099

u23e 2 0.003 4 0.004
Fuel carrier

Li’ A 282 0.027 282 0.019 492 0.027 492 0.019

F19 1,671 0.019 1,671 0.017 - 2,920 0.019 2,920 0.017
Moderator

Be’ 82 0.001 82 0.001 143 0.001 143 0.001

c'? 0.028 0.020 0.028 0.020
Fission products 3 0.006 6 0.007
Parasitic isotopes

U236 and others 6 0.017 13 0.020
Miscellaneous

Pa?33 (fuel) 2 0.010 3 0.011

Pa?33 (blanket)

Blanket carrier salt, core 0.052 0.048 0.052 0.047

vessel, and leakage

Neutron yield, 7 2.241 2.211 2.241 2.209
Total fuel inventory, kg 59 82 103 148
Cumulative net burnup, kg 0 ~59 0 -108
Net fuel requirement, kg 59 23 103 40
Regeneration ratio 1.114 1.061 1.114 1.055

9Neutrons absorbed per neutron absorbed by fissionable isotopes.

16

UNCLASSIFIED
ORNL-LR-DWG 38448

| b

CORE DIAMETER: 5 ft
LATTICE: TRIANGULAR, 8—in. CENTERS

142 FUEL: U233
FERTILE MATERIAL: Th232
9 \
[
<L
o 140 0 40
MO FUEL voy |
I_
<I
1
w
pd
L)
(&)
Lut
[n gy
1.06 0.15 FUEL VOLUME FRACTION, 125Mw | 7 —
} | T —— ]
0.15 FUEL VOLUME FRACTION, 250Mw |
]
0 2 4 6 8 10 12 14 16 18 20

REACTOR OPERATING TIME (yr)

Fig. 1.1.4. Regeneration Ratio vs Reactor Operating Time for Two-Region, Graphite-Moderated, Molten-

Salt-Fueled, Breeder Reactors.

17

UNCLASSIFIED
ORNL—-LR—-DWG 38149

120 ’
T
~Nw PLAN
\NVENTORY FOR 250
/ DOUBLING TIME, 19yr ~_ I
80 INVENTORY FOR 125-Mw PLANT Z i
o
S — < /
..—--—-——-- ) D“A !
a «?\'/ DOUBLING TIME, 22 yr
3 60 e | —
5 \fi?’ I //
@ pN’(
L PL |
g 40 o - i
C
g 433 pRODY / |
»
)
o
20
CORE DIAMETER: 5 ft
LATTICE: TRIANGULAR, 8-in. CENTERS
0 FUEL: U233 —
FERTILE MATERIAL: Th232
FUEL PROCESSING RATE: 12 TIMES PER YEAR
-20 -
0 2 4 6 8 10 12 14 16 18 20 22

REACTOR OPERATING TIME (yr) i

Fig. 1.1.5. Inventory and Excess U233 production vs Reactor Operating Time for Two-Region, Graphite-
Moderated, Molten-Salt-Fueled, Breeder Reactors with a Fuel Volume Fraction of 0.10.

18

220
200
180
160

440

100

FISSIONABLE ISOTOPES (kg)

=20

UNCLASSIFIED

ORNL-LR-DWG 38150

REACTOR OPERATING TIME (yr)

| | [ I

CORE DIAMETER: 5 ft

LATTICE: TRIANGULAR, 8-in. CENTERS

FUEL: U3

FERTILE MATERIAL: Th*>%

FUEL PROCESSING RATE: 12 TIMES PER YEAR

S
//WORY FOR 250-Mw PLANT
/-’—/
I INVENTORY FOR 125-Mw PLANT
| 1
| //
CUMULATIVE EXCESS U2 PRODUCTION IN 250-Mw PLANT | —"" |
/
—
/
-
" CUMULATIVE EXCESS U%** PRODUCTION IN

_________;_/ 125-Mw PLANT
2 4 6 8 10 12 14 16 18 20

Fig. 1.1.6. Inventory and Excess U233 Production vs Reactor Operating Time for Two-Region, Graphite-

Moderated, Molten-Salt-Fueled, Breeder Reactors with a Fuel Volume Fraction of 0.15.

19
was only =0.00077. It was concluded that Be9 is an appreciable poison

in homogeneous, molten-salt-fueled reactors, but has negligible effect

in graphite-moderated molten-salt-fueled reactors.

Oracle Code MSPR=-Cornpone 020

A revision has been maede of the element absorption edit of Oracle

code MSPR-Cornpone 020, which now permits an immediate edit on paper

tape. Previously it was necessary to wait for curve-plotter pictures.

It is now possible to compute the breeding ratio and prepare the input"

for Sorghum as soon as the case is run.

The cross-section tape has been cheecked and corrected where

necessary, and an edit has been obtained for the cross sections. The

elements now included in the cross section library are listed below:

Code Number

00
0l
02
03
Ol
05

20

Element
4233
4235
U238*

Thorium*
239

Fluorine

Oxygen

Boron
Sodium
Beryllium
Bismuth
Cerium

Two Fission Products

Deuterium

Iron

Element

Code Number

i

1

15
16

17
18

19
1A
1B

Nickel

Chromium

Molybdenum

Aluminum

Lithium (0.01 % Li
Carbon

Alloy-8
U23h

1l

Thorium (2)

Thorium (0)

Thorium (1)

Thorium (4)

Thorium (10)
Thorium (25)

Thorium (0.25)

6),

26 Hydrogen 2B Thorium (26G)

27 Thorium (0.75) 2C Thorium (29G)
28 Thorium (0.50) 30 Thorium (13)
29 Thorium (7) 31 PthO

- 2A Thorium (13G)

| o38%
There are 32 lethargy groups, with the last at 1180°F. Element UC38

has zero absorption cross sections in all groups. It must be combined
with one of the other U238 options listed, which have zeros for all
cross sections except absorption, to compose a complete set of cross

238. The options permit the selection of a set of

sections for U
absorption cross sections properly asdjusted with respect to resonance
saturation and Doppler broadening. The numbers in parentheses indicate
the concenfiration in mole % of UEBSFA in mixtures of LiI" and BeF2 such
as those used in the molten-salt reactors. Thorium is treated simi-
larly, and sbsorption cross section options for both mixtures with
lithium and beryllium fluorides and graphite (denoted by G in parenthe-

ses) are available.

Design of 30-Mw Experimental Reactor

Work is continuing on the 30-Mw, one-region, experimental, molten-
salt reactor described in the previous report. The basic reactor
concept has remained unchanged. More details of the equipment have
been developed, and various auxiliary system layouts are being studied.
A reactor site has been selected for design purposes, and perspective
drawings of a test reactor building have been prepared.

The bayonet-type fuel-to-coolant heat exchanger is being redesigned
in an effort to simplify the fabrication problems. Design data for use
in layout studies have been prepared for a U-tube, U-shell, coclant-

salt-to-steam superheater.

MR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p 3.

21

Fuel Drain System
The fuel drain system consists of four 2-1/2-ft-dia vessels

approximately 20 ft long. These tanks will be suspended in a vertical
position and manifolded to the reactor with pressure-siphon drain

lines. Two drain vessels will be required to contain the fuel system
inventory, and each pair of tanks will have a separate system to isolate
it from the reactor. The two extra drain tanks will provide capacity

for a spare salt volume, which may be used for fuel system cleaning or

decontamination operations.

The vessels will be preheated and afterheat will be removed with
a recirculating gas system. This gas system, as in the case of the
reactor gas system, will include a blower, heater, and cooler packages
enclosed in a loop.

The drain system criticality problem is being investigated. A
multiplication constant of less than 0.5 was obtained for one drain
vessel presumed to contain the entire fuel system UFh inventory in an
LiF-BeF,. salt mixture. As a further check, the U235

2
settle out in the vessel as an oxide compound, and multiplication

was assumed to

constants are being obtained for four different geometries of fuel

concentrate, with one edge reflected by the salt mixture. A maximum

multiplication constant has not yet been obtained, but, with the fuel | |
concentrated in a cube, the multiplication constant was 0.83. It is

believed that this value would increase slightly with further settling

of the fuel. The bottom section of the drain vessel will therefore be

made with a smeller diameter to reduce the potential hazard.

Gamma Heating in the Core Vessel of the 30-Mw Experimental Reactor

The heat generation in the core vessel of the experimental reactor

has been calculated using the Oracle program Ghimsr.s’6 The total heat

5L. G. Alexander and J. W. Miller, Heat Generation in the One-
Region, Experimental Molten-Salt Reactor, ORNL~2746 (to be published).

D. B. Grimes, Ghimsr, Gamma Heating in Molten-Salt Reactors,
ORNL CF 59-5-40 (to be published).

3

generation was 2.30 w/cm” at the inner surface of the core vessel.

The individual contributions to the total heating are given below:

Source | | - Heating (W/cms)
Prompt gemma rays - 0.61
Delayed gamma rays 0.52
Inelastic scattering gamme rays 0.34
Beryllium capture gamma rays 0.02
Fluorine capture gamma rays 0.07
Uranium capture gamma rays | | 0.15
Core vessel capture gemma rays 0.59

Total 2.30

1.2. COMPONENT DEVELOPMENT AND TESTING

Salt-Iubricated Bearings for Fuel Pumps

Hydrodynamic Journal Bearings

The sixth test of a journal bearing operating in molten-salt fuel
130 (LiF-BeF,
terminated on schedule. During this test the temperature of the salt

-UF), , 62-37-1 mole %), as described previously,l was

was increased in EOOF increments from 1200 to lBOOOF. Performance was
satisfactory throughout the entire test, which was conducted at a
constant journal speed of 1200 rpm and a constant radial load of 200 1b.
The total operating time for this test was 192 hr.

A seventh test was then performed with the bearing and journal used
in the fifth and sixth tests. This test also consisted of steady-state
operation at a journal speed of lQOO.rpm and a radial load of 200 1b,
but start-stop operations were carried out three times at each test
temperature. The temperature was varied as in the sixth test from 1200
to 1400°F in 50°F increments. On attempting a second restart at the
1400°F temperature level, an increase in the power required was noted,
and the test was halted. .The total operating time for this test was
216 hr. Inspection revealed a galled spot on the loaded side of the
bearing.

The eighth test was conducted to study the effect of increasing
the bearing radial clearance‘(measured at room temperature) to 0.007
in. All bearings previously tested had a radial clearance of 0.005 in.
Molten-salt fuel 130 at 1200°F was used for this test, and the bearing
radial load was varied from 50 to 500 1b at each of two journal speeds,
600 and 1200 rpm. After 356 hr of operation, including 39 start-stop
sequences, a sudden increase in power halted the test. At the time the
power increase occurred the journal was operating at 600 rpm, and the
load was being increased from 300 to 400 1b; the load had reached 382 1b.

From the standpoint of minimum film-thickness, these conditions were the

lMSR Quar, Prog. Rep. Jan. 31, 1959, ORNL-2684, p 39.

2L

most stringent yet imposed in these molten-salt~lubricated bearing
tests.

A carburized INOR-8 journsl was used with an uncarburized INOR-8
bearing for the ninth test. The tester was started and stopped 260
times during a period of operation of 272 hr at 1200°F. The jdurnal
- speed was 1200 rpm and the bearing was loaded to 200 1b. Both the
bearing and the journal were in good condition, with undamaged surfaces,
at the end of the test. The bearing 1s shown in Fig. 1.2.1 and the

Journal in Fig. 1.2.2.

Hydrodynamic Thrust Bearings

Assembly of the thrust-bearing tester described previouslyl is
nearly completed. The thrust-load actuator is being given preliminary

tests prior to being fitted to the rotary portion of the tester.

Test of Pump Equipped with One Salt-Lubricated Journal Bearing

Detailed design work is nearly complete on the modifications to be
made to a PK type of centrifugal sump pump in order to replace the
lower oil-lubricated bearing with a molten-salt-lubricated hydrodynamic
journal bearing. The external motor of the PK pump is being replaced
with an integral, totally enclosed motor. The configuration of the
modified pump is intended to simulate, in large measure, that of the
pump proposed for use in an experimental molten-salt reactor. Proposals
have been received for the fabrication of the motor and the pump shaft,
and purchase orders are being placed. An existing facility is being

prepared for tests of this pump.

Bearing Mountings

Means are being investigated for flexibly mounting INOR-8 bearings
to iNOR-B pump casings sSo that thermal distortiofi between the bearing
and the pump shaft journal can be accommodated. A diaphragm type of
mounting and a mounting that is somewhat similar to the Westinghouse

"Thermoflex" mount are being studied.

25

UNCL ASSIFIED
PHOTO 33924

9c

UNCLASSIFIED
PHOTO 33926

0] 1 2 3
! | | J
0 i 2 3 INCHES
| 1 | |
INCHES Fig. 1.2.2. Carburized INOR-8 Journal
After Operation for 272 hr at 1200°F
Fig. 1.2.1. INOR-8 Bearing After Operation for 272 hr at 1200°F with Molten-Salt Lubrication. with Molten-Salt Lubrication.

Mechanical Seals for Pumps

The modified Fulton-Sylphon bellows-mounted sea12 being subjected
to an endurance test in a PK-P type of centrifugal pump has accumulated
an additional 1848 hr of operation since the previous report period, for
a total of 12,478 hr. The pump has been operating at a temperature of
1200°F, a shaft speed of 2500 rpm, and a NeK flow rate of 1200 gpm. The
maximum test~seal leakage rate was 3.5 cmS/day; however, on an averaged
basis, the leakage is still negligible. Two pump stoppages occurred —
one for reworking-the drive motor commutator and replacing the motor

brushes, and one as a precautionary measure during a power outage.

Pump Endurance Testing

3

An MF type of centrifugal pump has continued in operation,~ and

has logged more than 15,500 hr (epproximately 1 3/4 yr) of continuous
operation. No maintenance has been performed on the pump during this
period. During the last 13,465 hr, a cavitation endurance test has been
under way with the pump operating at the steady-state conditions of 2700
rpm, 645 gpm, and 2.5-psig pump tenk cover gas pressure at 1200°F. During
the quarter the pump was stopped five times. Two stops were momentary
and were caused by power outages. One stop was of 65 min duration for
replacing brushes in the motor-generator set. One stop was of 30 min
duration for calibrating the pressure-measuring devices. One stop was

for 60 min for freeing the system throttle valve that had become stuck.

Frozen-Lead Pump Seal

The small frozen-lead pump seal being tested on a 3/16-in.-dia

shaft, as described previously, has operated continuously since it was

2MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p k2.

3MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p L2,

b, B. McDoneld, E. Storto, and J. L. Crowley, MSR Quar. Prog. Rep.
Oct. 31, 1958, ORNL-2626, p 23.

27

started on June 13, 1958. The accumulated operating time is more than
7500 hr. Excepfi fbr slight leakage of lead during the first 100 hr of
operation, there has been no further leakage. |

The large lead pump seal with a 3 1/4-in.-dia rotating shaft has
been operating for 3600 hr. Leskage of solid lead from the seal is
sporadic. The average leakage rate for the time operated is 9 cms/hr.
Operational data suggest that the seal should be redesigned to provide
better coolant control and to incorporate a short, resilient packing to
decrease the annulus between the seal and the shaft. A detailed
description of the equipment was presented previously.5 At the present

time, the operating'conditions are the following

Shaft speed 975-1000 rpm
Argon pressure over molten

lead in tank 0—-3 psizg
Seal cooling water flow rate 0.8 gpm
Temperature rise in cooling

o

water ~25°F

Temperature where seal is formed .l8O~19OOF

Techniques for Remote Maintenance of the Reactor System

Progress has been made in the design and construction, Fig. 1.2.3,
of a remote maintenance demonstration facility. The scheduled completion
date is June 30, 1959, and all phases of the work are on schedule.

The General Mills mechanical arm manipulator, 3 1/2- and 6-in.
Inconel piping subassemblies, dummy heat exchangers, salt dump tank,
dunmy reactor, two closed-circuit television systems (Fig. 1.2.4), and
the spool for Jjoining the pump and motor have been received. One of the
20 sets of freeze-flange pipe joints (Fig. 1.2.5) is on hand, and the
modifications required to adapt an existing PK type of pump for use in
this facility have been completed.

Eleven pipe preheater and insulation units that were designed to

MSR Quar. Prog. Rep. Jan. 31, 1950, ORNL-268%, p 43.

28

- ok um:mw%m
PHOTG 33887

‘ Fig. 1.2.3. Remote Maintenance Demonstration Facility Under Construction.

ot

Fig. 1.2.4. Closed-Circuit Stereo Television Equipment for Remote Maintenance Demonstration Facility.

UMCLASSIFIED
FHOTO 33885

i R R S UNCLASSIFIED
- e i : B PHOTO 33886 .

'
i -
I
i
: =

~ Fig. 1.2.5. A Freeze-Flange Joint for a é-in. Pipe in the Remote Maintenance Demonstration Facility.

31

contain electric heaters and insulation in one package are being fabri-
cated. These units are to be removeble and replaceable in short sections
with the use of the General Mills manipulator. Specially designed
spring-type pipe supports are being procured. Three gas and compressed-
air control cabinets are being fabricated.

Installation of the General Mills manipulator, heat exchangers,
dummy reactor, salt-circulating pump, salt piping, service piping, pump-

lube-oil piping, and electric heater control cabinets 1s in progress.

Design, Construction, and Operation of Materials
Testing Loops

Forced-Circulation Loops

The operation of long-term forced-circulation corrosion-testing
loops was continued. Fourteen test loops are presently in operation,
two Inconel loops were terminated as scheduled during the quarter, and
one new Inconel loop was started. A loop fabricated of INOR-8 is
presently being installed.

The two Inconel loops, designated 9344-2 and 9377-3 in Table 1.2.1,
were terminated after one year of operation. The facilities from which
these loops were removed have now been revised according to the latest
design6 for long-term operation. Of the 15 available loop facilities,
11 are of the improved design, and the remaining four loops are in
various stages of improvements.

Several momentary power failures caused interruption of power to
the loops during the quarter, but flow resumed in all the loops after
power was restored, and there were no freeze-ups. The power failures
that have occurred since the beginning of this molten-salt-reactor
corrosion-testing program and the results of such failures are described
in Table 1.2.2. '

The loops designated MSRP-12 and 9377-5 in Table 1.2.1 contain
molten-salt-sampling devices. It was necessary to add 600 g of salt

63. L. Crowley, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 36.

Table 1.2.1. For

" Composition

| A imat i
Loop Loop Material Number of pproximate  Approxim

Designation and Size Circulated Flow Rate Reynol. Comments
1o (gpm) Numbe
Fluid
9354-3 INOR-8 84 2.8 ration
Hot leg, % in. sched 40 4500
Cold leg, % in. OD, 5400
0.045 in. wall ‘
9344-2 Inconel, }'2 in. OD, 0.045 12 2.5 8200 Jan. 28, 1959 after one year of
in., wall
9377-3 Inconel, }'2 in. OD, 0.045 131 2 3400 ! March 18, 1959 after one year
in. wall on
9354-1  INOR-8, J, in. OD, 0.045 126 2.5 2000 ration
in, wall
19354-5  INOR-8, % in. OD, 0.035 130 1 2200 ration
in, wall
9354-4 INOR-8 130 2.5 y insert removed Feb, 17, 1959
Hot leg, 98 in. sched 40 3000} hr of operation; loop restarted
Cold leg, }'2 in. OD, ‘ 3500, fluid; normal operation other-
, 0.045 in. wall
MSRP-7  INOR-8, J in. OD, 0.045 133 1.8 3100 ration
in. wall
9377-4 Inconel, % in. OD, 0.045 130 1.8 2600 ration
in. wall
MSRP-6  INOR-8, %, in. OD, 0.045 134 1.8 2300 ration
in. wall
MSRP-8  INOR-8, J, in. OD, 0.045 124 2 4000 rqtion
in. wall
MSRP-9  INOR-8, J, in. OD, 0.045 134 1.8 2300 ration
in. wall ‘
MSRP-10  INOR-8, , in. OD, 0.045 135 2 3400 ration
in. wall o
MSRP-11  INOR-8, % in. OD, 0.045 123 2 3200 (ation
in. wall
MSRP-12  INOR-8, /, in. OD, 0.045 134 1.8 2300 cation
in. wall
9377-5  Inconel, J, in. OD, 0.045 134 1.8 2300 /qtion
in, wall
9377-6  Inconel, J in. OD, 0.045 133 1.8 3100 ation Feb. 26, 1959
in. wall
9Composition 12: NaF-KF-LiF (11.5-42-46.5 mole %) - Composition 130:
Composition 84: NaF-LiF-BeF, (27-35-38 mole %) Composition 131::
Composition 123: NaF-BeF y-UF ; (53-46-1 mole %) Composition 133:,
Composition 124: NaF-BeF ,-ThF 4 (58-35-7 mole %) Composition 134:
Composition 126: LiF-BeF ,-UF ; (53-46-1 mole %) Composition 135:.

bJ, L. Crowley, MSR Quar. Prog. Rep. Jan. 31, 1958, ORNL-2474, p 31.
Ibid., p 32.

d). L. Crowley, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 36, Fig. 1
®J. L. Crowley, A Sampling Device for Molten Salt Systems, ORNL-2688 (to be

33

Table 1.2.2. Summary of Power Faillures That Have Occurred During
Operation of Forced-Circulation Corrosion~Testing Loops

Number of Number of  Number of Number of Loops
Date Loops Loops Loops in That Failed as
Operating  Affected Which Salt  Result of Power
Froze Failure
December 17, 1957 5 5 5 1
April 6, 1958% 9 6 6 3
August 5, 1958 7 7 éb None
November T, 1958 1k 14 None None
January 15, 1959 15 15 None None
March 15, 1959 15 12 None None

®putomatic controls installed on most locops between April and August.

bThese loops did not have automatic controls; only part of the salt
froze.

to replenish the inventory and thus raise the fluid level to the normal
height in loop 9377-5 after 27 samples had been taken. The molten salt
was introduced into the pump bowl through the pump flange without
interruption of the flow or operation of the loop. Analyses of the
samples removed from these loops periodically are presented in

Section 2.2.

In-Pile Loops

The in-pile loop which was operated in the MIR during the previous
quarter was disassembled, and the cause of the partially plugged purge
line, which resulted in the release of activity, was determined. Oil
which leaked past the pump shaft seal and filled the oil trap on the
purge outlet line was polymerized by radiation, and the polymerized oil
plugged the outlet line. Chemical and metallographic analyses of the
fuel and container materials have not been completed.

The second in-pile loop was completed and inserted during the

MTR shutdown of April 27. Modifications were made to the internal

35

purge system to minimize the probability of purge line plugging. In
addition, changes were made to the external purge system which should
eliminate the possibility of a repetition of the activity release.
These changes included the installation of a charcoal trap in the
purge inlet line to prevent back-diffusion of fission gas into un-
shielded lines. A prototype in-pile pump identical to that installed
in the second loop has accumulated over 3600 hr of satisfactory

operation.

36

1.3. ENGINEERING RESEARCH

Physical Property Measurements

Enthalpy and Heat Capacity

Determinations were made of the enthélpies, heat capacities, and
heats of fusion of three additional beryllium-containing fluoride salt
mixtures: salt 133 (LiF-BeF2~ThFh, 71-16-13 mole %), salt 134 (LiF-
BeF ,-ThF), -UF) , 62-36.5-1-0.5 mole %), and salt 136 (LiF-BeFe—UFu, 70-
10-20 mole %). The results of these measurements are presented in
Table 1.3.1 as the constants a, b, and ¢ appearing in the correlating
equations,

2
HT - HBOOC = a + bT + T

and

c._=Db+ 2T ,
p

where H is the enthalpy (cal/g); Cy the heat capacity (cal/g.°C); and

T, the temperature (OC). The experimental enthalpy data for these three
mixtures, along with earlier_resultsl for related mixtures, are summarized
in Fig. 1.3.1. It may be seen by comparing the results for salts 133

and 136 with the curve for mixture 130 (LiF-BeFe-UFu, 62-37-1 mole %)

that altering the salt composition toward larger percentages of the
high-molecular-weight fluorides (ThFu and UFA) causes a substantial
decrease in both the enthalpy and heat capacity (on a unit weight basis).
Replacement of the LiF in salt 126 (LiF-BeFQ-UFh, 53=46-1 mole %) with

NeF (mixture 123; NaF-BeF,-UF), 53-46-1 mole %) results in similar,

though smaller, reductions in the enthalpy and heat capacity. This

effect will be studied further in order to establish a general correlation
for the prediction of enthalpies and heat capacities in the LiF-BeFe—
UFM-ThFZ system.

lW. D. Powers and R. H. Nimmo, MSR Quar. Prog. Rep. Oct. 31, 1958,
ORNL-2626, p uh.

37

o3

Table 1.3.1.

Enthalpy Equation Coefficients and Heats of Fusion of
Several Fused Salt Mixtures
Enthalpy Equation Heat of Fusion (cal/g)
Salt Mixture Phase = Temperature Coefficients® Temperature - HS
Range (°C) a b c (oc)
x 1077
LiF-BeF ~ThF) -UF), Solid 100-350 -b.9  +0.22k  +3k4.65 450 6L
(62-36.5-1-0.5 mole %) Liquid 500-800 -2.1  +0.545 -4.03
LiF-BeF,~ThF) | Solid 100-400 -7.7 +0.209 +4.58 500 h
(71-16-13 mole %) Liquid 550-800 b7 404730 -11.90
LiF-BeF ~UF), Solid 100-400 -5.6 +0.153 +3.50 450 53
(70-10-20 mole %) Liquid 550-800 -17.8 +0.289 -3.21

SThe enthalpy is given as HT - H3

¢ =b + 2¢T.
P

= a + bT + CTQ; the heat capacity is then evaluated as
Viscosity

The viscosities of salt mixtures 133 and 136 have been obtained
over the temperature range from 550 to 800°C with the "skirted" capillary
efflux viscometer previously déscribed.2 The viscometer cups used for
the measurements were those employed in the earlier studies with salts
130 and 134%. Recalibration of these cups with the NaNOZ-NaNO?)-KNO3

(40-7-53 wt %) mixture showed negligible changes in the kinematic

viscosity—efflux time relation. The results, correlated in the form

where u is the viscosity (centipoise), T is the temperature (%K), and A
and B are experimentally determined constants, are as follows:
Salt 133 A = 0.0526; B = L4838
, Salt 136 A = 0.0489; B = L4847
The date are plotted in Fig. 1.3.2 and compared with data from
measurements of salts 123, 126, 130, and l3h.3 The "high" results with

i

i

mixtures 123 and 126 will be rechecked with the new type of viscometer.
In Fig. 1.3.3, the data for salt 136 are compared with Mound Laboratory
results for the same mixture that were obtained with a Margules design
viscometer consisting of two concentric Inconel cylinders. The outer

of the two cylinders contalned the fused salt and was rotated at a fixed
speed with reference to the inner stationary cylinder. The viscosity
was determined from the torque exerted on the inner cylinder. The
reason for the discrepancy between the two sets of data has not yet been

established.

Surface Tension

The surface tension of molten-salt mixture 134 was experimentally

determined using the maximum-bubble-pressure technique.3 The results,

°MSR Quer. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 65.
3W. D. Powers, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 38.

l;B. C. Blanke et al., Density and Viscosity of Fused Mixtures of
Lithium, Beryllium, and Uranium Fluorides, MIM-1086 (March 23, 1959).

39
1, VISCOSITY (centipoise)

450 500 550 600 650 700 750 800 825
100
s — | e
/I//p
90 CR
60 \3 %o 47‘(//.?
~ 65
LQUIDUS >~
~TEMPERATURE (°
\\
PN
\\\ \
\\ \
10— S~ —
SALT COMPOSITION (mole %) e ———— N
NO.  LiF NoF BeFp ThFs UFg e N
8 — =
126 53 46 f — T~
. 123 53 46 1 \\s \\
133 7 16 13 7 \\\
136 70 10 20 BSOS \
~
4l— 134 82 365 10 05 BN
130 62 37 f TN
2_ | N
00014 00013 0.0012 0.0014 0.0010 00009

40

UNCLASSIFIED
ORNL-LR-DWG 38454

[ T
L!QUIDUS TEMPERATURE (°C)

COMPOSITION {mole o)

LiF NaF Befp, Thigy UF4

62 37 i 134
62 365 1 05 /A 126 Pl
53 46 ! / /
53 46 f P } 33
74 s 13 -
70 10 20 T
i
/ —
/

13024134
—

T

—— SALT NO. 130

500

400
_ SALT
o
< NO.
g
° 130
= 126
= 123
& 133
o 136
£ 200 —
i2]
T
i
<

100

o}
0

Fig. 1.3.1.

400 500
TEMPERATURE (°C)

TEMPERATURE (°C)

600 700 800

Enthalpy-Tempetature Relations for Several Beryllium-Containing Salt Mixtures.

UNCLASSIFIED
ORNL-LR-DWG 38152

Fig. 1.3.2. Viscosity-Temperature Relations for Several Beryllium-Containing Salt Mixtures.

given in Fig. 1.3.4, can be represented by the equation,
o (dynes/cm) = 272.2 - 0.143 T (°cC),

to within +5% over the temperature range from 500 to 800°C. The salt
134 data are essentially identical with the results obtained for salt

130, except for a slight difference in the temperature dependence.

Heat~Transfer Studies

Flow calibration of the system designed for determihing surface
film formation in molten-salt systems by heat transfer coefficient
measurements has been completed. A Tygon-coupled mockup was used, and
water was the working fluid. The test unit is shown in Fig. 1.3.5, and
the key components are indicated. The over-all system pressure drop
was obtained with a 100-in. mercury manometer connected at the pump inlet
and outlet lines. The data indicate a pressure difference of about 48 psig
for a flow of 2.8 gpm. This corresponds to a test-section Reynolds
modulus of 11,000 at 1200°F with a salt such as mixture 130 (LiF-BeF,
UF), 5 62-37-1 mole %). Initial attempts to determine the flow rate vs

pump speed characteristic by a weight method were unsuccessful, and a
rotameter was used in the final calibration. A magnetic-pulse pickup
coupled with a Hewlett-Packard counter yielded the pump speed to within
+1 rpm over any lO-sec counting period. The characteristic curve obtained
was very nearly linear between 2000 and 6000 rpm, ranging from 1 to 3
gpm, respectively.

Design work is continuing on the system support structure to ensure
against buckling of the test sections due to thermal expansion and to
prevent electrical short-circuiting of the test-section heating currents.

Welding of system components is in progress.

L

VISCOSITY (centipoise)

UNCLASSIFIED
ORNL—LR-DWG 38453
TEMPERATURE (°C)

500 550 600 650 700 750 800

% 7 T [T T T ]
AN

REANN
N

s

N

O  MOUND LABORATORY

L

1.4 1.3 1.2 1.1 1.0 0.9 x1073
1/T K}

Fig. 1.3.3. Comparison of Independent Measurements
of the Viscosity of Salt Mixture 136 (LlF-Ber-UF4,
70-10-20 mole %).

UNCLASSIFIED
ORNL-LR~-DWG 38454

250 }
SALT 434\

—. 200 = e |

£ \'-7-_=.§

< SALT 130 * —— o 130
) o~
= 150 !
g 134
.9

=

jun]

}_

w 100

Q

E

o

2

1]

b 50

0
400 450 500 550 600 650 700 750 800 850

TEMPERATURE (°C)

Fig. 1.3.4. Surface Tension of LiF-Ber-ThF4-UF4 (62-36.5-1-0.5 mole %).

Lo

UNCLASSIFIED
PHOTO 464147

HEAT
EXCHANGE

€4

Fig. 1.3.5. Flow Calibration Mockup of Molten-Salt Heat Transfer System.

Hydrodynamic Studies

The pressure drop through the sintered-metal-filled annulus of &
double-walled tube has been determined.5 A tube of this construction
has been proposed for use in steam-generation systems associated with
molten-salt power reactors. It is hoped to provide good thermal contact
between the low-pressure primary salt and the high-pressure steam and
at fihe same time effectively isolate the two fluids in the event of s
wall failure. The sintered medium would be filled with static helium,
at an intermediate pressure, which would give an indication of a bresk
in either wall. |

The results of this study of the flow of helium through two speci-
mens of double-walled Inconel tubing containing sintered Inconel of
different porosities are given in Fig. 1.3.6. The correlating lines

can be expressed by the equations:
| 0.979
AP
= 0.342 ( L)
for specimen 1, with a porosity of approximately 59%, and

0.960
G = 0.0046 (FAIM)
g L

for specimen 2, with a porosity of approximately 25%. In these corre-
lating expressions, G is the mass velocity based on the geometric
cross-sectional area (1b /ft3), P is the mean helium den81ty (1p /ft3),
AP is the pressure drop through the sintered material (1b /ft ), and L
is the length of the sintered section. The experimental precision was
estimated to be +8%. The two curves in Fig. 1.3.6 are nearly parallel,
with slopes only slightly less than unity, and are in accord with the

statement of Darcy's law for the flow of gases in porous media,

°J. L. Wantland, H. W. Hoffmen, and R. L. Miller, Flow Through
Sintered Inconel Annuli, ORNL CF 58-9-36 (Sept. 17, 1953).

6
As opposed to an estimated pore void area at any cross section.

Ll

UNCLASSIFIED
ORNL—LR-DWG 32943 A

100
(]
&
A
50 &
SPECIMEN NO. {; 0®
POROSITY, ~ 59 % g8y —
o
| /
&
X
20 y
9
o o®
/t‘ ’“)
—~ ©
N y. e
- /
£ ® o2
£ o
.DE ,// ®
- - P
> SPECIMEN NO. 2 4
£ 0 POROSITY,v25%, ~ ¢®
o [
.| ,'
L)
> “.
a 7
a yd
2' "
S 2 /
°/
/| P
1 L
/
v
L 4
0.5
0.2
04 7
to 100 1000 5000
_ _ 6
PAP/L by 1o, [17)
Fig. 1.3.6. Flow vs Pressure Drop Characteristics for Two Annuli Filled with Sintered Inconel.

hs

If it is assumed that the data follow Darcy's law exactly, the specific
permeability coefficient, Bo’ can be obtained for the two samples by
using the given porosities and mid-range vaelues of G, and PAP/L. In

-12 .2

this manner, B_ was calculated to be ~62 x 10 £1° for specimen 1 and

o
~164 x 10-14 £t2 for specimen 2. These results can be compared with B_
values of 216 x 10~1° £t° for an unconsolidated send of 37% porosity and
68 x lO"lh

porosity. In addition, the slope of approximately unity indicates that

£t° for consolidated small alundum particles with a 24%

the flow is in the laminar regime.

L6

1.4, INSTRUMENTS AND CONTROLS

Molten-Salt-Fuel lLevel Indicators

Visual and x-ray examination of a level probe used in the tests
described previouslyl have been completed. The probe appears to be in
excellent condition, both mechanically and electrically. Disassembly
of the probe is in progress, and specimens have been sent to the
metallurgy department for corrosion examination.

'Two new Inconel "I'"-tube-type level elements have been completed
and installed in test vessels. Two identical level-measuring systems
are now complete, and prefilling checkouts of the systems are in progress.

Fabrication of an INOR-8 level element and test vessel is approxi-'

mately 80% complete.

INOR-8 High-Temperature Pressure Transmitters

An order was placed with the Taylor Instrument Company for six
INOR-8 pressure transmitters and indicating systems. The units are
scheduled for delivery in early June.

The transmitter body will be fabricated from INOR-8 plate. The
diaphragm material will be 0.005-in.-thick INOR-8 shim stock. The
manufacturer believes on the basis of preliminary tests that satisfactory
diaphragms can be made of this material. No unusual difficulties are
expected in the fabrication of these units.

The six units will be composed of three having a range of O to 50
psig, and three with a range of O to 100 psig. All units are of the

pneumatic-indicator type.

1R. F. Hyland, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p L47.

L7

PART 2.

MATERIALS STUDIES
2.1. METALLURGY

Dynamic Corrosion Studies

Corrosion studies were completed of four Inconel and four INOR-8
thermal-convection loops, and 12 thermal-convection-loop tests were
initiated. The operating conditions for the new loops are given in

Table 2.1.1. Two forced-~circulation loops have completed the scheduled

Table 2.1.1. Operating Conditions for New Thermal
Convection-Loop Tests

Composition Number  Maximum Fluid-Metal  Scheduled

Loop of Salt Being Interface Operating
No. Material Circulated® Temperature (OF) Period
1236  Inconel 134 1250 1 yr
1237 Inconel 134 1350 1 yr
1238  INOR-3 134 1350 1yr
1239  Inconel 133 1350 1 yr
1240  INOR-8 133 | 1350 1yr
1247 INOR-8 136 1250 1000 hr
12k2  INOR-8 136 1350 1000 hr
1243 Inconel 135 1250 1000 hr
124  INOR-8 135 1250 1 yr
1245  Inconel 135 1250 1yr
1246 INOR-8 135 1350 1lyr

1247  Inconel 135 1350 1l yr

fComposition 134: LiF-BeF ~ThF) -UF) (62-36.5-1-0.5 mole %).
Composition 133: LiF-BeF,-ThF) (71-16-13 mole %).
Composition 136: LiF-BeF -UFM4(70-10-20 mole %).
Composition 135: NaF-BeF,-ThF) -UF) (53-45.5-1-0.5 mole %).

N

oM

51

test program but have not yet been examined, and one new forced-
circulation loop test was started. The present status of all forced-
circulation-loop tests now in progress is given in Chapter 1.2 of

this report.

INOR-8 Thermal-Convection Loops
Metallographic examination of INOR-8 loop 1185, which circulated
salt 126 (LiF-BeFe-UFh, 53-46-1 mole %) for one year at a hot-leg

temperature of 12500F, revealed no observable attack in either the
hot- or cold-leg sections. A photomicrograph of a typical hot-leg
section from this loop is shown in Fig. 2.1.1. Further, no evidence
of mass transfer was found in examinations of cold-leg sections or
analyses of after-test salt samples. Results of salt analyses

(presented in Table 2.1.2) indicate the level of metallic impurities

Table 2.1.2. Analysis of Salt Mixture 126 Before and After
Circulation in Loop 1185 for One Year

Major Constituents Minor Constituents
(3t %) | (ppm)
Sample Taken U Be Ni Cr Fe
Before test 6.53 10.1 30 40 235
After test | 6.4o 10.6 35 140 250

in after-test salt samples to be effectively the same as in before-
test samples.

Metalloéraphic examination results for the INOR-8 loops 1227,
1228, and 1229 are summarized in Table 2.1.3. These three loops were
each operated for 1000 hr, and the maximum fluid-metal interface
temperature was 12500F. Metallographic examinations of the hot and
cold legs of these loops showed no evidences of surface defects of
the types normally associated with fluoride salt attack; however,

well-developed intergranular cracks from 0.5 to 2 mils deep appeared

52

€S

Table 2.1.3. Results of Metallographic Examinations of Inconel and INOR-8
Thermal Convection Loops
Meximum Fluid-
Metal Interface Metallographic Results
Loop Test Tempgrature Salt Hot-Leg Cold-lLeg
No. Material  Period (°F) No.? Appearance Appearance
1182 Inconel 1 yr 1350 126 Heavy intergranular voids Surface pitting 0.5
to a depth of 15 mils mil deep
1188 Inconel 1 yr 1250 8k Heavy intergranular voids Grain-boundary
to a depth of 9 mils penetrations to a
depth of 1 mil
1189 Inconel 1yr 1250 130 Heavy intergranular voids  Grain-boundary
to a depth of 7 mils penetrations to a
, depth of 0.5 mil
1230 Inconel 1000 hr 1350 130 Moderate intergranular Few penetrations
voids to a depth of 4 < 1 mil deep
mils
1227 INOR-8 1000 hr 1250 134 No attack No attack
1228  INOR-8 1000 hr 1250 133 No attack; fabrication No attack
flaws found in both legs
1229  INOR-8 1000 hr 1250 135 Same as loop 1228 Same as loop 1228

qComposition 126:
Composition 8k:
Composition 130:
Composition 13k4:
Composition 133:
Composition 135:

LiF-BeF,-UF), (53-46-1 mole %).

LiF-NaF-BeFs (35-27-38 mole %).

LiF-BeF,-UF), (62-37-1 mole %).
LiF-BeFo-ThF) -UF), (62-36.5-1-0.5 mole %).

LiF-BeF,-ThF), (71-16-13 mole %).

NeF-BeF 5-ThF), -UF), (53-45.5-1-0.5 mole %).
at random intervals along the exposed surfaces of all three loops.

An examination of as-received specimens of tubing used for these loops
was subsequently made, and similar cracks were detected in these
specimens. A typical flaw is shown in Fig. 2.1.2. The flaws were
apparently not detected Dby x-ray inspection of the tubing before loop
fabrication. Therefore ultrasonic inspection of loop tubing is now

being utilized to supplement existing inspection practices.

Inconel Thermal-Convection Loops

Three of the four Inconel loops for which metallographic exami-
nation results have been recently obtained were operated for one year,
while the fourth was operated 1000 hr. The examination results are
summarized in Table 2.1.3. As indicated, the hot legs of all the
Inconel loops operated for one year showed heavy intergranular void
formation. Of particular interest is loop 1182, which circulated salt
126 for one year at 13500F. Examination of this loop revealed heavy
intergranular void formation to a depth of 15 mils, as may be seen in
Fig. 2.1.3. Chemical analyses of the salt showed no significant
changes in the impurity content. However, examination of the trap
area, which was the coldest part of the loop, revealed a large quantity

of metallic particles. These particles are being analyzed.

General Corrosion Studies

Penetration of Graphite by Molten Fluoride Salts

The problems associated with the use of unclad graphite as a
moderator in a molten-salt reactor system are being studied. The
possibility exists that fuel salts would enter the pore spaces in the
graphite and create problems associated with (1) reactor fuel inventory,
(2) hot spots, (3) effects from fission-gas release associated with
fuel in the pores, and (4) effects of thermasl expansion of the salt
as a result of thermal cycling of the reactor. The large volume change
which fluoride salts exhibit on melting could cause cracking or spalling

of the graphite if sufficient pressure were built up by salt entrapped

54

Fig. 2.1.1. Specimen Taken from Hot Leg of INOR-8
Thermal-Convection Loop 1185 at Point of Maximum

| ed )
UNCLASSIFIED | X~ RN B UNCLASSIFIED
T-16554 e ' ' : 4 T-16822

007

008

Q08

010

Reik]

012

013

Loop Temperature (1250°F). Loop was operated for 1229.
1 year with salt mixture LiF-BeF,-UF, (53-46-1

mole %).

UNCLASSIFIED fl
T-16577 T

Q

<z

007

.08

009

010

Ol

Rely-

013

Fig. 2.1.3. Specimen Taken from Hot Leg of Incone! Thermal-
Convection Loop 1182 at Point of Maximum Loop Temperature
(1350°F). Loop was operated for 1 year with salt mixture
LiF-Ber--UF4 (53-46-1 mole %).

009

010

Ol

012

013

Dié

Fig. 2.1.2. Specimenof INOR-8 As-Received Tubing
Used in Thermal-Convection Loops 1227, 1228, and

55

in pores. ‘ 4

An experimental investigation has been startéd to determine under |
what conditions salts will penetrate graphite and to ascertain the
physical damage that might occur in the graphite as a result of thermal
cyeling. Both static-pressure and dynamic-pressure tests are being
conducted.

Static-pressure penetration tests have been made with graphite-
molten salt systems at 150 psia and l3OOOF. In preparation for these
tests, graphite specimens were partially degassed by heating at 23729F
for 5 hr in a vacuum of <4 u and then cooled to room temperature in
a pure argon atmosphere. When assembling the apparatus, the graphite
was exposed.to room atmosphere for approximately 5 hr. The sapparatus
was then evacuated to < 12 # and held at 5720F for 15 hr to remove
lightly adsorbed contaminants.

In test, the graphite specimens were completely submerged in molten
salts at l3OOOF, and a 150-psi pressure was applied for 100 hr by using
an argon blanket gas. After the tests the specimens were examined for
penetrations of salt into the graphite by using radiographic, metallo-
graphic, and/or chemical‘analysis techniques. The specimens were also
welghed to detect weight changes.

Both TSF and CON graphite vere tested individuelly in LiF-BeF-UF)
(62-37-1 mole %, fuel 130). Macroscopic examination of radiographs
of these graphite specimens did not reveal any penetration by fuel 130.
There was essentially no weight gain (<0.01%) of the TSF graphite
specimen, but there was a weight gain of approximately 1% for the CCN
graphite specimen. Metallographic examinations and chemical analyses
are being made to determine the location of the fuel 130 and/or reaction
products in the CCN graphite. The TSF grade will be exemined in like
manner.

A static~pressure penetration test at 13OOOF was also made with
fuel mixture 30 (NaF-ZrFu-UFu, 50-46-4 mole %) and CCN-grade graphite
to assist in evaluating testing techniques. Radiographic examination of

the graphite specimen used‘in this test indicated that the fuel had

56

penetrated it throughout. The weight gain of the specimen indicafied
that 29% of the calculated pore volume of the graphite had been filled.
Additional static-pressure penetration tests are planfied using fuel
mixture 130 with TSF, AGOT, and ATJ grades of graphite to evaluate

the effects of pressure and temperature on penetration.

A dynamic~pressure penetration test appératus has been devised
which allows the movement of fuel around the graphite specimen to wash
away reaction products that might form at the graphite surface and thus
plug pore spaces. This test equipment, which is further discussed
later in this chapter under "Welding and Brazing Studies, " is designed
to apply pressures up to 350 psia on the fuel side of the graphite
while the opposite side is being evacuated.

The CCN-graphite specimen that was penetrated by fuel 30, as
mentioned above, was thermally cycled to determine whether thermal
expansion of the fuel 30 in the graphite pore spaces would cause
cracking or spalling of the graphite. During the thermal cycle, thermo-
couples indicated temperature changes from room temperature to 1436°F
near the outer surface and room temperature to 779OF at the central
axis of the specimen within a 3-min period. In the first two minutes
of the cooling part of the cycle, the 1436°F outer surface temperature
was lowered to 842°F and the 779°F central temperature was lowered to
752OF. Radiographic and macroscopic examinations of the graphite
specimen at room temperature did not revesl any damage to it.

Since the volume expansions of fuels 30 and 130 from room temper-
ature to their melting temperatures are 17% and 11%, respectively, the
lack of damage to the graphite suggests that some penetration of molten
fluoride salts into the graphite pore spaces might be tolerated. The
specimen will be subjected to additional thermal cycles, and other
penetration and thermal-cycling tests of various types of graphite and

fluoride salts will be made.

57

Uranium Precipitation from Molten Fluoride Salts in Contact with

Graphite
The fuel mixture LiF-BeF,-UF) (62-37-1 mole %) was reported

previouslyl to have precipitated U02 in tests in which the fuel mixture

was contained in graphite crucibles at l3OOOF. The graphite-salt

systems were tested in evacuated Inconel containers, and the precipitate
was observed by using radiographic techniques. In control tests under
similar conditions with two different batches of fuel mixtures in Inconel
containers without graphite crucibles, no UO2 precipitate was detected
by radiographic examination.

To assist in determining the role of graphite in causing uranium
precipitation from graphite-fuel systems, a set of five tests was run
with different quantities of graphite and different salt-to-graphite
contact areas. The tests were made concurrently at l3OOOF, and the
fuel mixture used in each test was from the same batch. Graphite,
grade CCN, was used that had been machined and cleaned by sonic means
in ethyl alcohol.

The test setups used are illustrated in Fig. 2.1.4, and the tests
are described below. In test A, the fuel was tested in the absence of
graphite in order to test both the quality of the fuel and the assembly
methods used in fabricating all the test equipment. In test B, the
fuel was in contact with a large area of a large quantity of graphite
in order to establish a reference for uranium precipitation from the
fuel; the dimensions and configuration of the test duplicated those of
the standard graphite—fuel compatibility tests. In test C, the fuel
was tested in the presence of, but not in direct contact with, a large
quantity of graphite to determine whether the uranium precipitation
was caused by degassing of the graphite. In test D, the fuel was tested
in contact with a small area of a large quantity of graphite to determine,
in conjunction with the results of test B, whether the contact surface-
to-volume ratio for the graphite and fuel affected the quantity of

uranium precipitation from the fuel. In test E, the fuel was tested

Lvisr Quar. Prog. Rep. Jan. 31, 1959, ORNL-2634, p 80.

58

UNCLASSIFIED
ORNL-LR-DWG 37260

w

2]

)

T

=

o

T

Yo

DIU

S Q

A

Q =

w O

1 = O A

w r = g

LANUPRL
w

WogR T o

>

L 206 F 0

Y FE _

n

(&

=

E
,il|||1|l||t|i|
N—_———"_=————————__—

N RN/
/7 : 3 V%
,, \\\ 3 //fl////// ////\\\\
/7 =

SN S5

7 e e ————————\ /7

Nt — N\

\ Nyl — — Q) ,/6, / /
W ///N/W/, /% = /,, N S =
b e e S

Fig. 2.1.4. Test Setups Used to Study the Precipitation of uo, from Fuel 130 in Graphite Crucibles.

59

in contact with a small area of a small quantity of graphite to
determine, in conjunction with the results of test D, whether the
quantity of graphite affected the quantity of uranium precipitation
from the fuel.

Radiographic examinations of each of the test setups at total
accumulated test times of 5, 10, and 100 hr were made. No uranium
precipitation was observed for the control, test A (Fig. 2.1.k). A
small amount of uranium precipitation occurred in test E. A moderate
and approximately equal amount of uranium precipitation was found in
each of the other three tests. These results support the conclusion
arrived at previously2 that the uranium precipitation observed in
fuel 130 was the result of the fuel reacting with oxygen supplied by
degassing of the graphite.

Tests are in progress to determine whether a suitable flush can
degas the graphite to such a degree that it will contain fuel 130

without causing uranium to precipitate from the fuel.

Thermal-Convection-Loop Tests of Brazing Alloys in Fuel 130
A scheduled 1000-hr test of Inconel and INOR-8 lap joints brazed

with various alloys and exposed to fuel 130 in the hot leg of a thermal-
convection loop has been completed. Similar tests scheduled for 5,000-
and 10,000-hr periods are continuing. The configuration of these
thermal-convection loops and the manner in which the brazing alloys

are incorporated in them was described previously.3 The brazing alloys

tested were:

Alloy Composition
Coast Metals No. 52 89% Ni—5% Si-4% B—2% Fe
Coast Metals No. 53 81% Ni-8% Cr—4% Si—4% B-3% Fe
General Electric No. 81 70% Ni—20% Cr—10% Si
Gold—Nickel Alloy 82% Au—-18% Ni
Copper 100% Cu
W. H. Cook, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 62. :
3E. E. Hoffman and D. H., Jansen, MSR Quar. Prog. Rep. June 30, 1958
ORNL-2551, p 62. .

60

The temperature of the circulating salt in the region of the test
s?ecimens was l3OOOF.

Metallographic examination showed that all the brazing alloys had
good flowability on both Inconel and INOR-8 base metals. There was a
tendency for the formation of diffusion voids in the fillets of the
joints brazed with the gold—nickel alloy. The General Electric No. 81
alloy was heavily attacked on Inconel base material. Coast Metals Nos.
52 and 53 were depleted at the fillet surface to a depth of 1 to 2 mils,
Some slight cracking of the alloys occurred at the brazing alloy-base
metal interface. Pure copper showed good corrosion resistance and no

cracking.

Thermal-Convection-Loop Tests of the Compatibility of INOR-3, Graphite,
and Fuel 130 :

Fuel mixture 130 (LiF-BéF2-UFu, 62-37-1 mole %) was circulated in
an INOR-8 thermsl-convection loop containing a 10-in. tube of TSF

graphite in the hot leg in order to ascertain whether the graphite would
cause carburization of the INOR-8. The loop operated for 4000 hr with
a hot-leg temperature of 1300°F. |

Metallographic examination of a loop section adjacent to the
graphite insert revealed that the carbide precipitates in the grain
boundaries were slightly larger near the salt-metal surface than
throughout the rest of the specimen (Fig. 2.1.5); however, similar
precipitates have been observed in INOR-8 tubing from corrosion loop
tests with no graphite present.

Microhardness measurements of the area 2 mils from the salt-metal
surface were of the same magnitude as those observed in the rest of
the specimen. These values ranged randomly from 160 to 180 DPH.

A carbon content of 0.037% was found by chemical analyses of
two successive 5-mil layers taken at the inner surface of the INOR-8
tubing. An additional series of layers to a depth of 35 mils was
analyzed and found to contain a range of carbon values from 0.01% to
0.026%. Consequently, these different values are not considered to

be significant, especially in view of the large quantities of inclusions

61

003

04

UNMCLASSIFIED
Y -29049

0

UNCLASSIFIED

62

'Ber'
te cap-

iF

ion Loop that Circulated L

Convect

8 Specimen Taken from a Thermal

fuel 130) for 4000 hr at a Hot-Zone Temperature of 1300°F

The (a) Outer and (&) Inner Surfaces of an INOR

Fig. 2.1.5
UF, (62-37

jacent to a graphi

ion ad

is from a sect

imen

The spec

-1 mole %,
in the hot leg.

aqua regid.

Etchant

sule incorporated

in this heat of material, as shown in Fig. 2.1.6.

Mechanical Properties of INOR-8

Creep Tests

Most of the creep tests presently being conducted in fused salts
are low-stress, long-time tests. The results of these tests will pro-
vide accurate data in terms of stress, total strain, and creep rates
for INOR-8 at times in excess of 10,000 hr. None of the test specimens
failed during the past quarter and therefore no data are available to

report. These experiments are being run at 1100 and 1200°F,

Fatigue Studies
The fatigue properties of INOR-8 at 1100 and 1500°F are being
studied at Battelle Memorial Institute under a subcontract from the

Metallurgy Division. A rotating-beam-type of test is used at
frequencies of 100 and 3000 rpm. The initial results reported were
for the 100 rpm and lBOOOF conditions. The data are summarized below:

Stress (psi) Cycles to Failure
45 x 107 8 x 103

40 | 34 x 10

35 160 x 103

30 730 x 10°

30 930 x 10°

27 72 x 10°

(discontinued)

Thus, the stress to produce fatigue failure in 1 x 106 cycles is
slightly over 29,500 psi. Under similar test conditions, the stress
to produce fatigue failure of Inconel in 1 x 106 cycles is about 18,000
psi.
Shrinkage Characteristics of INOR-8

A certain peculiarity in the behavior of INOR-8 has beeh noted

almost since the start of the current testing program. In creep tests,

63

6L

B

e

&

aiaey
P

INCHES

CLASSIFIED
“¥-29052

e

Fig. 2.1.6. Cross Section of INOR-8 Tube that was Adjacent to Graphite
Insert

in 4000-hr Thermal-Convection Loop Test.

there is an initial period during which virtually no creep occurs. This
is followed by normal creep. In relaxation tests, the load must be
increased during the early stages in order to maintain a constant strain.
These observations are important, since at low stresses the effects are
manifest over a thousand or more hours and the results in terms of creep
rate and total strain vs time measurements are attenuated.

The behavior pattern suggested that the metal might be contracting
as a result of metallurgical instabilities. Isothermal dilatometry
measurements indicate that there is contraction in certain heats of
INOR-8 but not in others, notably SP-16 and SP-19. Metallographic
examinations showed that those specimens which contracted contained
many large areas of second phase material. The stable heats had equi-
axed grain structures with fine precipitates concentrated mainly in the
grain boundaries. There appear to be no significant differences in
either the major or minor alloying constituents in the heats examined,
except for the carbon content. However, stable heats have been found
that range from a special low-carbon melt up to heats containing carbon
in excess of 0.1%. The specimens which contracted had carbon contents
which fell within this range. Currently the contraction phenomenon is
being studied by means of resistivity measurements to determine temper-
asture-dependence and reaction rates. This study will be continued, since

all heats have shown a plateau in the creep and relaxation values.

Materials Fabrication Studies

Effect on INOR-8 of Aging at High Temperatures
The experiments which were being run to determine whether INOR-8

exhibits a tendency to embrittle in the temperature range between 1000
and 1400°F have been completed. ©Specimens which had been aged for
10,000 hr have been tensile tested, and no significant differences from
annealed specimens have been found. The data from these tests are

presented in Table 2.1.k.

65

99

Table 2.1.Lk,

from Heat SP-19 (0.06% C)

Effect of Aging Upon the Tensile Properties of INOR-8 Specimens

Specimen Annealed 1 hr at 2100°F

Specimen Annealed and Aged 10,000 hr

at Test Temperature

Test - - - - -
Tbmfg;§ture gi?:;;ih éii:igth El°?§§ti°n Temfiiifiiure gifi:;éih si;:igth El°?§§ti°n

(psi) (psi) (°F) (psi) (psi) ’
Room 114,400 Lk, 700 50 1000 117,700 45,100 50
1100 120,000 47,500 49
1200 116,000 46,800 46
1300 115,500 45,800 43
1400 115,000 43,600 40
1000 93,000 28,300 L6 1000 97,700 L6
1100 93,000 28,900 50 1100 90, 500 Lo
1200 82,400 27,500 37 1200 81,900 32
1300 69,900 28,000 2l 1300 76,100 2l
1400 61,800 26, 200 21 1400 65,100 20

It is concluded from this group'of tests and from the previous
tests'?D of specimens aged for 500, 1000, 2000, and 5000 hr that INOR-8
does not exhibit any embrittling tendencies that can be attributed to
high-temperature instability.

Triplex Heat Exchanger Tubing

Work is continuing on the development of techniques for fabricating
a heat exchanger tube made up of two concentric tubes with a porous
annulus between to permit the passage of a gas for leak detection. The
materials of construction presently being investigated are a porous
nickel core clad on the outer and inner surfaces with Inconel or INOR-8.
Two methods of incorporating the porous core into the annular space of
the concentric tubes are.being considered: first, tamping loose nickel

powder into the annulus and developing a porous core bonded to the

- outer and inner tubes by suitable drawing and sintering operations; and,

second, bonding a prefabricated porous core to the outer and inner tubes
by suitable drawing and sintering operations. Results obtained by the
first method were described previously. Because of difficulties
encountered with an oxide layer being formed at the core-to-cladding
interface and causing poor bonding, this method of fabrication has
been discontinued in favor of the second method.

Porous nickel sheet is available from Micro Metallic Corp. in

four common grades, which differ in mean pore opening, as follows:

Grade Mean Pore Opening (in.)
E 0.0015
F | 0.0008
G ' 0.000k4
H 0.0002

ILH. Inouye, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 67.
°H. Inouye, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 67.
6MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p 85.

67

The sheet material has a density of approximately 50 to 60% of
theoretical, and 1/16-in.-thick stock can be rolled into a 3/Wk-in.-OD
tube. Twenty-five feet of both grades E and G rolled into tubular
shape have been ordered for cladding studies.

An attempt is presently being made to clad with Inconel a small,
sample, porous nickel strip received from Micro Metallic. The as-
received strip was annealed for 1 hr at 1800°F and then bent around
the outside of a 0.625-in.-0D Inconel tube. The porous nature of the
resulting ring can be seen in Fig. 2.1.7, which shows an enlarged
view of the surface. Channeled spacers were tack-welded to the Inconel
tube adjacent to the porous ring to prevent slippage of the ring
during drawing and to provide a means of filling the annular space.

The assembly is shown in Fig. 2.1.8. An Inconel tube was placed over
the assembly and the resulting triplex assembly was cold drawn from
1.000 in. OD x 0.500 in. ID to 0.855 in. OD x 0.468 in. ID. The tri-
plex aséembly is now being preparéd for sintering to promote bonding

of the porous ring to the Inconel cladding. A vacuum will be maintained
in the annulus of the triplex assembly during sintering. An evaluation
of the porosity and bond between ring and cladding will subsequently

be made.

Similar cladding experiments are to be performed by Superior
Tube Co. on Subcontract No. 1112. It is anticipated that these results
will help determine the feasibility of fabrication of a triplex tube

by a commercial vendor.

Welding and Brazing Studies

Procedures for Welding INOR-8
The fabrication of INOR-8 components for the wide variety of MSR

applications requires that suitable welding procedures be developed
for material ranging in size from thin-walled tubing to heavy plate.
An inert-arc welding specification, RMWS-12 (ref 7) is available for

"R. M. Evens (comp. and ed.), Reactor Material Specifications,
TID-7017, p 195 (Oct. 29, 1958). .

63

Fig. 2.1.7. Surface of Prefabricated Porous Nickel Strip Received from
Bending Around a 0.625-in.-OD Inconel Tube. 7]/2X.

INCHES
1 2

I I I |

{

Micro Metallic Corp. After

UNCLASSIFIED

Y-29097

Fig. 2.1.8. Assembly for Cladding Porous Nickel Ring with Inconel. Channeled spacers provided for

restraining ring, facilitating evacuation of assembly, and filling annular space.

69
INOR-8 material up to 0.100 in. thick, and additional procedures are
being qualified for thicker sections in accordance with the methods
prescribed by the ASME Boiler Cod.e.8

Several lengths of 5-in. sched-4O pipe have been fabricated from
l/h-in.-thick plate, and a welding procedure will be developed which
will be applicable over the thickness range 1/8 to 1/2 in. A welding
procedure for l/2~in. plate will also be developed, which will be
applicable for thicknesses up to 1 in. The tests to be conducted
include reduced-section tension, side-bend, face-bend, and root-bend
tests, as well as radiographic and metallographic examinations.

It is expected that these welding procedures will be extremely
useful in fabricating containment vessels and test components from
INOR-8. These procedures will serve also as an essential supplement
to the mechanical property data which are being accumulated for the
presentation of INOR-3 as a case before the ASME Boiler and Pressure

Vessel Code Committee.

- Mechanical Properties of INOR-8 Welds
Studies are continuing in an effort to improve the high-temperature

ductility of INOR-8 weld metal. These include the investigation of
techniques to deoxidize and purify weld filler metal during the casting

of original ingots. One promising method involves the use of a vacuum- |
melted basic charge with additions of small quantities of aluminum,
titanium, manganese, silicon, boron, and magnesium. Seversl different
heats of weld metal éontaining these and other additions have been cast
and fabricated into weld wire. Analyses of these heats are presented
in Table: 2.1.5, and analyses of commercial materials are included

for comparison. The mechanical property values obtained to date for
these materials are summarized in Table 2.1.6, and properties of other
materials ére included for comparison. The results of the mechanical
property tests on these filler metals will permit an evaluation of the

relative merits of the various deoxidation and purification practices.

BASME Boiler and Pressure Vessel Code, Section IX (1956).

70

T.

Table 2.1.5.
Containing Alloying Additions

72% Ni—16% Mo—-4% Fe-8% Cr

Basic melt charge:

Analyses of INOR-8 Weld Metal Heats

Carbon .
Heat Content Analysis Additive (wt %)
Number (wt %) Source Mn Si Al i R Mg
Hastelloy W 0.03 ORNL 0.58 0.07 0.15 0.08 0.02
(25% Mo—T7% Fe—5% Cr—
2.5% Co—bal Ni)
Westinghouse M-5 0.08 Vendor 0.79 0.19 0.02
Haynes SP-19 0.06 Vendor 0.48 0.0k
ORNIL, MP-3 0.06 Intended 0.50 0.10 0.20 0.20 0.005 0.03
ORNL 0.36 0.25 0.25 0.20 0.00k 0.02
ORNL MP-k 0.06 Intended 0.50 0.10 0.20 0.20 0.03
ORNL 0.3+ 0.05 0.20 0.24 0.002

cl

Table 2.1.6.

Mechanical Property Studies on INOR-8 Weld Metal

Tensile Strength (psi)

Flongation (% in 1 in.)

Heat At Room At Room

Number Temperature At 1200°F At 1500°F Temperature At 1200°F At 1500°F
Hastelloy W 127,000 94,800 70,400 37 33 25
Westinghouse M-5 118,900 4, 000 55,800 36 18 5
Haynes SP-19 115,900 70,400 53,000 38 18 10
ORNL MP-3 108,000 67,800 49, 200 46 o5 13
ORNL MP-4 107,600 68,000 50,100 45 22 9

Fabrication of Apparatus for Testing the Compatibility of Molten Salts
and Graphite

Equipment was fabricated which will be used by the General Corrosion
Group to study molten salt penetration of graphite in a dynamic, high-
pressure system. Before this apparatus could be fabricated, it was
necessary to develop a‘method of attaching an Inconel tube to a hollow
cylindrical graphite specimen and thereby forming a leaktight connection.
This was accomplished by brazing with a commercially available braze
alloy composed of silver, titanium, and copper that was found to wet
vacuum-degassed graphite.

The large difference in thermal expansion of the graphite and the
‘brazing alloy caused some shear cracking in the graphite. However, |
metallographic sectioning of sample joints indicated that the cracks
were rather limited in length, and pressure testing of.the completed
joint with air indicated no leakage other than that associated with
the inherent porosity of the base graphite. |

The graphite-to-Inconel assembly prior to brazing is showvn in
Fig. 2.1.9, and the completed specimen is shown in Fig. 2.1.10. The
completed Inconel test rig with associated entry, drain, and purge lines

is showvn in Fig. 2.1.11.

3

UNCLASS‘ FIED
Y.28696

Fig. 2.1.9. Grcp‘nite-to-\ncone\ Assemb\y Befote Brazing

UNCLASS\F!ED
Y.28695

Fig. 2.1.11% Comp\e'ted inconel

Test Rig:

Fig. 2.1.10. Graphife-io-lncone\ Assembly Afrer Brozing.

2.2. CHEMISTRY AND RADIATION DAMAGE

Phase Equilibrium Studies

The System LiF-BeF _-ThF,
A revised phase diagram for the system LiF-BeFE-ThFu is presented

in Fig. 2.2.1 that includes data obtained by increasing the equilibration
period in thermal-gradient quenching experiments to three weeks. Quenched
samples from such experiments revealed that the area of single-phase
ternary solid solutions involVing 3LiF~ThFh is greater than previously
reported;l’2 it occupies approximately the triangle indicated by cross
hatching in Fig. 2.2.1. Invariant equilibria in the system LiF»BeFE-ThFh
are listed in Table 2.2.1. The limits of the area of the single-phase

Teble 2.2.1. Invariant Equilibria in the System LiF«BeFQ-IhFu

Composition (mole %) Invariant Type of Solids Present at
LiF BeF2 ThFh Temperature  Equilibrium Invariant Point
(ec)
17 81 2 Lot Peritectic ~ ThFj, LiF-4ThF),
» and BeFp
33.5 6b 2.5 451 Peritectic  LiF.A4ThF), LiF-2ThF),
\ and BeF2
L7 51.5 1.5 355 Eutectic 2LiF.BeF,, LiF.2ThF),,
and BeF,
60.5 36.5 3 431 Peritectic LiF-2ThF), 3LiF-ThF)
(ss), and 2LiF-BeF,
66 2.5 4.5 448 Peritectic LiF, 2LiF-BeF,, and
3LiF.ThF) (ss)
63 29 8 452 Peritectic  3LiF.ThF), (ss),
TLiF-6ThF),, and
LiF. 2ThF),

1R. E. Thoms et al., MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551,

2R. E. Thoma et al., MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626,
P 9.

75

UNCLASSIFIED
ORNL-LR-DWG 37420

Thi, 11H4

TEMPERATURE IN °C
COMPOSITION IN mole %o

LiF-4ThE,
7LiF-6ThE
T A LiF-2The,
3LiF~ThF4 sS
448 LiF-2ThE,
‘ P 897 A
2LiF-BeF,
[N
7TLIF-6ThE
N \
P 597
P 565 \
3LiF-ThE,
FAN \‘
SORS
S, e
LiF AN
845 2LiF-BeF2/5<T)of45o 400 400 450 500

P 465 £ 370

Fig. 2.2.1. The System LiF-BeF,-ThF,.

76

3LiF'ThFu solid solution are:

" Composition (mole %)

F B I,
50 25
58 16 26
58 21 21

The revised diagram shows that the composition LiF-BeF,-ThF) (67.5~
17.5-15 mole %; liquidus temperature, 500°C) has a higher ThF) concentra-
o= ThF),
(71-16~13 mole %); also the ligquidus temperature is slightly lower. The

tion than breeding mixtures of current interest, such as LiF-BeF

occurrence of the unusual ternary solid solution in this system provides
these breeding mixtures with an advantage; on cooling, the first solids
and last liquid to freeze are much closer in composition that would be
the case if no solid solution occurred. A description of the crystalli-
zation process for the composition LiF—BeFe-ThFA (67.5-17.5=15 mole %)
demonstrates this point. On cooling to 500°C, the two solid phases
3LiF-ThF) (composition located in triangular solid solution area) and
7LiF-6EhF4 coprecipitate on a boundary curve. As these two solids
continue to freeze upon further cooling, the equilibrium liquid decreases
in thorium concentration and leaves the boundary curve before reaching
the perifiactic point at LiF-Bng-THFu (63-29.5-7.5 mole %, 452°C). The
liquid then changes composition toward another peritectic at LiF-BeFE-
ThF), (60.5-36.5=3 mole %) and freezes completely in the viecinity of the
latter composition. Since the quantity of liquid remaining in the region
of the peritectic is much less than if no solid solution existed, most

of the solids formed on cooling IiF—BeFe-ThFh (67.5-17.5=15 mole %) will
be formed between 500 and hSOOC, and tendencies toward segregation will

be reduced.

The System NaF-BeF  -ThF,

Evidence has continued to accumulate on the extensive substitution

of UFM for ThFu to give solid solutions in systems containing BeF2 and

alkali metal fluorides. The resulting solid solutions are compounds of

7

alkali metal fuorides with UFA and ThFh; they melt at considerably lower
temperatures than the corresponding thorium compounds. Accordingly, the
phase diagrams of ternary systems of an alkali fluoride with BeF2 and
ThFh are useful as guides to selections of converter-reactor fuel mixtures
from the otherwise extremely complex quaternary systems containing UFM‘
Progress toward completion of the phase diagram for the system NaF-
BeFE-ThFu is shown in Fig. 2.2.2. The composition of the single ternary
compound occurring in the system and the location of the boundary curve
separating the primary phase fields of ThFh and this ternary compound
have not been determined. The compositions that show the maximum solubility
of THF), at 550°C are NeF-BeF -ThF) (76-10-1% mole %) and NeF-BeF,~ThF)
(50-38.5-11.5 mole %).

The System NaF‘ThFL"UEg

The identity ;nd approximate locations of the primary phases in the
system NaF-ThFh-UFu have been determined. Because of the solubility of
UFA in compounds such as hNaF-ThFu, 2NaF-ThFu, 3N&F-2THFA, and 7NaF-6ThFu,
there are no phase fields for pure binary thorium compounds in the ternary

diagram.

The System SnFE—NHIHF2

The SnFE-NHuHFE system was investigated because of its potentiali=-
ties as a strongly oxidizing, low-melting solvent for reprocessing fuels.
Liquidus temperatures were noted by visually observing precipitation
upon cooling. At temperatures highexr than 15000, the solution bubbled
and formed a scum which not only interfered with observations but also
changed the composition of the melt. Therefore reliable data were
obtained only in the range O to 40% SnFE. The compound NHMHF2 melts at
125°C, and a 15 mole % addition of SnF,
about 100°C. TFurther additions of SnF2 increase the liquidus temper-
ature to 150°C for the mixture containing 40 mole % SnF

lowers the melting point to

o
Heating a mixture of 35 mole % NHL}HF2 and 65 mole % SnF2 resulted
in considerable evolution of gas and gave a melt with a liquidus temper-

ature of 2&000, which is well above the melting point of SnF2 (2180C).

78

UNCLASSIFIED
ORNL—LR—DWG 38155

ThFg . TEMPERATURE IN °C
‘ COMPQOSITION IN.mole %

111
) £ = EUTECTIC
P = PERITECTIC
PRIMARY PHASE FIELDS
A = NafF
= 2NaF - Th,
= 2NaF * BeF,
NqF-ZThF4 n
P 83 \

= NaF - Thfy
' P 730 ‘

£ 705 A
3Nof - 2Th,

= 3NaF - 2ThF,
£ 690 A
2NaF - ThF,
. £ 618
P 645 4 \
N ' \
4NGF - ThF, ‘ ‘

= NaF : 2Thf
= NaF - Bef,
= TERNARY COMPOUND

C - IO Mmoo o0

3 £ 530
AN /\/ \ BeF,

Fig. 2.2.2. The System NaF-BeF ,-ThF .

79

|
|
The solidified product gave a simple but unidentified x-ray diffraction
pattern. Thus it appears that the sample was predominantly a new
compound. Petrographic results confirmed the presence of a new conpound
and showed that still another unidentified phase was present to the

extent of about 10%.

Solubility of PuF in Converter Fuels

Interest in the possibility of converting U 238 to the fissionable

Pu239 in molten-fluoride-salt fuels led to measurements of the solu-~
bility of PuF3 in the mixture LiF-BeF,-UF) (70-10-20 mole %). The
apparatus and techniques were described previously. 3 For the first
solubility determlnatlon, weighed amounts of LiF, BeF UF&’ and PuF3
were melted directly in the filtration apparatus and heated to a maximum
temperature of 716°C in a mixed atmosphere of argon and HF. The mlxture
was then cooled and equilibrated at 562 to 567°C for 105 min before
filtering at 567°C. Three other filtrates were obtained from a melt
that was pretreated by heating LiF, BeF,, and UF) in e platinum dish,
with NHhF-HF added to prevent hydrolysis, at a maximum temperature of
72500 for 30 min and cooling in an argon atmosphere. A plot of the

data in Table 2.2.2 as the log of molar concentration of PuF. versus

the reciprocal of the absolute temperature indicated that thz solu-
bility value obtained at 567°C was about 0.15 mole % too high to Tit
the straight line through the points for pretrested melts. The reason
for this slight discrepancy is not known. All the PuF3 solubilities
obtained in LiF-BeF -UF) (70-10-20 mole %) are higher than those
obtained with LiF-BeF2 mixtures having about the same LiF concentration

Separation of Li7F from LiTF-BeF
According to the LiF-BeF

2
~NaF ternary diagramh the addition of

2

3C J. Barton, W. R. Grimes, and R. A. Strehlow, Solubility and
Stability of PuF; in Fused Alkali Fluoride-Beryllium Fluoride Mixtures,
ORNL~-2530 (June 11, 1953).

R." E. Thoma, Phase Diagrams of Nuclear Reactor Materials,
ORNL-2548 (to be published).

80

Table 2.2.2. Solubility of PuFz in LiF-BeF,-UF)
(70-10-20 mole %)

Filtration
Temperature Plutonium in Filtrate
(°c) | Pu (wt %) PuF, (mole %)
558 3.43 1.27
567 - k.ot 1.52
597 L.s7 1.70
658 6.50 2,48

NaF to LiF--BeF2 (63-37 mole %) places the resulting composition in the
primary phase field of LiF, where a decrease in temperature causes
precipitation of LiF; under favorable conditions about 70% of the LiF
should separate as solid. An initial experiment has been performed in
which NeF was added to an LiF-BeF, mixture at 700°C and cooled to 490°C.
Analytical results on filtrates at 600°C and 490°C showed that only

23% of +the contained LiF remained in the solution at the lower temper-

ature.

Fission-Product Behavior

Precipitation of Sm‘F3 with CeF3

Previously reported studies

5,6 of the solubility of rare earth

fluorides in molten salts have shown that undesirable fission products,
such as SmF3, can be precipitated as solid solutions by CeF3 additions
and proper temperature adjustment. The removal is effected by the

exchangerr between the melt (d) and the solid solution (s):

W. T. Ward, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 88.
6MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268L4, p 100.

7The equilibrium constant for this exchange is independent of
temperature; however, the total rare earth solubility shows an expo-
nential variance with 1/T.

81

SmF3 (a) + Ce:F3 (s)e== CeF3 (a) + SmF3 (s) (1)
Experiments have shown that the exchange
> ¥
Ce*F3 (a) + LaF3 (s)= LaF3 (a) + Ce F3 (s) (2) _

is an analogue of exchange (1). Thus Ce*F3 (1abeled), which has
desirable tracer characteristics, can be used in exchange (2) to
approximate the behavior of SmF3 in exchange (1).

A proposed method of salt purification utilizing exchange (1)

would be to pass the salt through an isothermal bed of solid Ce}?3 to

lower the S.m’F3 content and then to lower the temperature of the effluent

salt to decrease the total rare earth content. To determine the
feasibility of the method, information is required on the rate of
exchange. The first exploratory rate test was conducted at 5OOOC with

an agitated CeF3-LiF-BeF2 melt (1500 ppm CeF3, 62 mole % LiF, 38 mole
% BeFE) to which an excess of LaF3 was added. Within 1 min, the CeF

content was decreased to 400 ppm; within 5 min the system was at

3

equilibrium (300 ppm CeFB) with respect to exchange (2). The second
test employed a 1lk-in. horizontal column of 3/k-in. tubing packed with
100 g of +18 mesh LaF.. The packed colum was charged at 300°C with
LiF-BeF,, (62-38 mole %) containing 1000 ppm CeFB, and 12 g of melt,
not counting the holdup volume, was transferred through the LaF. at

3
18 gpm with a pressure drop of 3 psi. The CeF. content of the effluent

ranged from 80 to 30 ppm. Attempts to continuz the experiment at a )
later time with the same packing gave evidence that the liquid was

bypassing the LalF3 (s) and that future tests with vertical columns

should give better performance.

Chemical Reactions of Oxides with Fluorides in Molten-Fluoride-Salt i
Solvents

The chemical reactions of oxides with UFM in molten fluoride
solvents have indicated that uranium can be separated from fission

products by fractional oxide precipitation. As mentioned previously,

8J. H. Shaffer, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 92.

82

BeO and water vapor appear to be desirable precipitating agents because
they do not introduce extraneous constituents into BeF2 mixtures.
Investigations have continued on the removal of uranium from solution
by reaction of UFH with BeO in a column-extraction process and on the

fractional precipitation of uranium with water vapor.

Extraction of Uranium with BeO. The reaction of UF), with BeO to
produce UO2 has been studied in several experiments.8’9 Observations
of the reaction in a packed column show several interesting characteris-
tiecs. A eutectic mixture of LiF-KF (50-50 mole %) containing 1 mole %
UFh was twice passed unldirectionally through a small column packed
with extruded BeO (1/16 in. in diameter and 1/8 in. long) at 600°C.
The column contained approximately three times the amount of Be(
o+ During the first pass, 34 .6%

of the uranium was removed from 2 kg of liquid mixture at an average

required for converting the UFA to U0

flow rate of 700 gpm under a pressure drop of 5 psi. A second pass
removed an additional 30% of the initial uranium from solution. How-
ever, the average flow rate during the second extraction was consider-
ably slower (30 gpm), since a pressure drop of approximately 15 psi
developed. A third extraction was not attempted.

Calculations based on the experimental data show that 16% of the
beryllium oxide in the column reacted with UF&‘ It is interesting to
note, however, that only 18% of the reacted beryllium was present in
the column effluent at the completion of the experiment. A similar
retention of Bng,_probably on the surface of the BeO pellets, was
previocusly noted in the conversion of Zth to ZrOe.

In future experiments, the effectiveness of BeO beds for removing
lower concentrations of uranium from molten fluoride mixtures will be

investigated.

Precipitation with Water Vapor. The effective removal of uranium
as UO, from an LiF-BeF, (63-37 mole %) solvent by the reaction at 600°C

2

95. H. Shaffer, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551,

83

of water vapor with UFA has been demonstrated.B In reprocessing

molten fluoride reactor fuels, however, the separation of uranium

from the fission-product rare earths would be a primary requisite.

An experiment has shown that the reaction of water vapor with CeF3

in LiF-BeF, (63-37 mole %) is too slow for detection by a fairly
sensitive radiochemical tracer technique (precision, <:ip.2%). In

two subsequent experiments, water vapor was allowed to react at 600°C
with UF) (1 mole %) in LiF-BeF, (63-37 mole %) containing 1.25 wt %
CeF3. In each case, UO2 was formed without any detectable precipitation

of Ce203 or BeO. The urenium concentration in scolution at the stoichi-
ometric point of the reaction was on the order of 100 ppm in each
experiment.

The sharpness of this separation is very favorsble from the
standpoint of reprocessing schemes, and it points to a method for
removing uranium without using fluorine. It also leads to a method
for removing uranium and thorium simultaneously that would permit the
rare earths to be selectively precipitated in another step that could
be followed by redissolution of the uranium and the thorium in
reclaimed barren solvent.

An experiment for studying the reaction of water wvapor with'UFu
dissolved in LiF-KF (50-50 mole %) indicated a much lower rate of
reaction as a result of strong complexing of both 0  and Uh+ ions
in the melt. A negligible quantity of HF was evolvéd, and x-ray |
diffraction examinations of filtrate samples taken from the reaction
mixture showed the presence of KF-2H20. Treatment with 8500 meq of
water resulted in the precipitation of 64 meq of uranium from solution
and the evolution of 30 meq of HF. A comparison with the reaction of
water vapor with UF, dissolved in LiF-NaF (60-40 mole %) is planned

as a further study.

8l

Gas Solubilities in Molten Fluoride Salts

Solubility of Neon in LiF-—BeF2
The investigation of the solubility of neon in a mixture of LiF-BeF2
(64-36 mole %) has been completed at 500, 600, 700, and 800°C covering

the pressure range of O to 2 atm. As in &ll previous studies of the

solubility of noble gases in molten fluoride salts, Henry's law is
Obeyed, and the solubility increases with temperature. The solubility
decreases with increasing atomic weight of gas, and the enthalpy of
solution increases.

Henry's law constants for this system are (3.09 + 0.09) x 10-8,
(k.63 + 0.01) x 10‘8, (6.80 + 0.09) x 10*8, and (9.01 + 0.15) x 107
moles of He per e of melt per atm at 500, 600, 700, and SOOOC, respec-
tively. A heat of solution of 5880 cal/mole was calculated from the

temperature dependence of these constants.

Solubility of CO, in Na'F-BeF2
The solubility of CO, in molten NaF-BeF, (57-43 mole %) has been
examined at 500, 600, 700, and 800°C at pressures from O to 2 atm. The

experimental data illustrate the linear dependence of solubility with
saturating pressures. Henry's law constants, K, at these temperatures
are (7.93 + 0.36) x 10'8, (7.05 + 0.1k4) x 10'8, (8.32 + 0.08) x 10"8
moles of CO2 per.cms of melt per atm. Of all the gas solubility data
for molten salt systems obtained to date, these eare the first to show
a nonlinear log X versus 1/T plot. A minimum is observed on this plot
in the vieinity of 600°C.

-Preliminéry measurements of 002 solubility in the eutectic mixture
LiF-NaF-KF (46.5-11.5-42 mole %) were questionable. It appears that
the presence of dissolved oxygen-containing species increases the
solubility of 002;
lously purified solvent before and after adding known quantities of a

This will be checked by measurements in a meticu-

soluble oxide.

85

Chemistry of the Corrosion Process

Samples from Operating Loops

Periodic sampling of the melts in forced-circulation corrosion-
test loops, as described in the previous report,lo has continued. One
loop is fabricated from INOR-8 and the other from Inconel. Both loops
were charged with the selt composition LiF-BeF-ThF)-UF), (62-36.5-1;0.5
mole %) containing about 400 ppm of chromium, and the maximum wall
temperature is 1300°F. |

The chromium concentration in the INOR-8 loop (MSRP No. 12) reached
a plateau at about 550 ppm after about 1200 hr of opération and has
maintained this level for the last 1000 hr. In the Inconel loop
(9377-5) the chromium éoncentration in the salt rose gradusally during
the first 400 hr of operation to a value of 550 ppm. The increase
was more rapid than in the INOR-8 loop but was uniform. However, after
the first 400 hr, the chromium concentration began increasing very

rapidly, as indicated below:

Chromium
Time Concentration
(hr) (ppm)
500 800
650 | 950 ’
800 1100
1700 2000 )

Since sample removals decreased the quantity of salt in the loop,
more salt was added after 2250 hr to permit further monitoring. Samples
taken before and after the addition of 0.6 kg of fresh salt showed 2100
and 1900 ppm of chromium, respectively. The loop was allowed to circu-
late for three days (72 hr) and again sampled, and 2050 ppm of chromium
was found; after 2600 hr, 2350 ppm of chromium was present. The latter

value may represent an approach to a steady-state level, but the chromium

19MsR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268L4, p 106. )

86

concentration is consideréblyvhigher than was expected in the absence
of oxidizing impurities.

Radioactive Tracer Analyses for Iron in Molten Fluoride Salts

Most investigations of corrosion behavior depend on accurate
analyses at low dilution of Fe't and other structural metal ions. For
example, the validity of chromium diffusion experiments in systems
consisting of a molten salt and a chromium alloy depends on the purity

of the salt with respect to NiF, and FeF,.,. Both compounds form CrF

2 2 2

from chromium; for example,

FeF,, + CrCe— CrF, + e .

Experimental equilibrium constants show that the equilibrium concen-
tration of FeF2 in the presence of chromium should be too low to detect.
However, analyses obtained in the diffusion experiments indicated that
FeF2 was present at concentrations ranging from 100 to 200 ppm in
NeF-ZrF), (53-47 mole %) which had been in contact with Inconel ([Cro]z
0.16 wt fraction). This anomaly suggested that finely divided iron
might be passing through the sample stick filters under certain con-

ditions. To investigate this possibility, 800 ppm of labeled FeF,. was

dissolved in a nickel container filled with NaF-ZrF) (53-47 mole ;)
solvent and then completely removed from solution by reduction with
zireconium. Standard counting procedures verified that no labeled Fe++
or iron was present in filtered portions of the melt after the zirconium
addition, although an average wet analysis of 205 ppm of iron was
reported for the same material. It seems definite that currently used
procedures result in misleading results from wet chemical analysis and
that the presence of iron in the melt as sampled is not the explanation.

The discrepancy is being investigated.

Activities in Metal Alloys

The measurements of thermodynamic activities for nickel in the

Ni~NiO system, as obtained from an electrode concentration cell with

a molten electrolyte, have been delayed by a series of experiments

87

designed to test the approach to equilibrium obtained by various
annealing procedures for the electrodes. These experiments have not
given definite results. Another series of experiments has established

that the results are not being invalidated by thermoelectric potentials.

Vapor Pressures of Molten Salts

The lowering of the vapor pressure of CsF (vapor préssure, 83 mm
Hg) at 1000°C by 20 mole % additions of alkaline earth fluorides has
been measured as part of a study of the effect of cation size and charge
on the thermodynamic properties of fluoride melts. The smallest cation
Mg++ (radius, 0.783) gives a lowering of 30 mm Hg; ca’ (radius, 1.063),
26 mm Hg; and Ba (radius, 1.433), 18 mm Hg, following the expected
order. Freezing-point depressions that reflect the same effects are

also being measured.

Permeability of Graphite by Molten Fluoride Salts

Permeability tests on various types of reactor-grade graphite with
molten fluoride salts have continued. In the routine procedure for
impregnating graphite with a molten salt, the graphite samples were
degassed under vacuum at 900 to 950°C and then treated with a molten
salt, such as 1iF~MgF2‘(67.5—32.5 mole %), while still held under
vacuum. After the graphite samples were completely covered with salt
the vacuum was relieved and a pressure of 15 psig of helium was applied
to the system for 48 hr. At the end of 48 hr, the pressure was
relieved, and the molten salt was transferred away from the graphite.

In previous tests, graphite rods 3-in. long and 1/4, 1/2, and 1 in.
in diameter had been used. For routine testing, l-in.-dia rods have
been adopted.

By using welded container vessels, shortening vacuum lines, and
using extreme care in assembly, the efficiency of impregnation has been
increased, as indicated by an average weight gain of 12.5% compared with

8.5% in earlier runs on reactor-grade TSF and AGOT graphite samples.

88

The objectives of the current exploratory tests are to determine
(1) how effective salt impregnation of graphite is in preventing or
decreasing penetration of possible molten-salt reactor fuels at the
maximum reactor temperature, (2) the degree of penetration into untreated
graphite of typical reactor fuels under forced impregnation (vacuunm
and pressure) conditions at the maximum reactor temperature (1250°F),
(3) the degree of penetration of typical reactor fuels into untreated
graphite under normal reactor operating conditions (in a 1000-hr test
in a circulating-salt system), and (4) the distribution of the main
fuel components (uranium, thorium, and beryllium) in the graphite when
penetration occurs.

Samples of three special types of graphite were obtained from the
National Carbon Company for testing. These were identified as follows:
ATT-82, ATL-82, and AGOT-82. The number 82 after each type refers to
a special process used by National Carbon to make impervious graphite.
Because of the geometry of the available samples, the rods were machined
to 3/h in. OD and 3 in. long. Attempts to impreghate these rods of
special grephite with LiF-MgF2 in the usual manner gave the following
results. The ATJ-82 graphite showed a loss in weight of 0.9%. The
ATL-82 graphite showed a loss in weight of about 0.1%. The AGOT-82
graphite showed a gain of 2.0%, and samples of TSF graphite, included
for comparison, gained 5.5%.

The rods were then subjected to a 1000-hr soaking test at 12500F
under a static helium pressure of 1 psig in stirred LiF-BeF,-UF) (62-
37-1 mole %). The ATJ-82 gained 0.2% in weight, the ATL-82 gained
0.4%, the AGOT-82 gained 2.1%, and the TSF gained 2.8%. These gains
are based on the weights of the rods before and after the 1000-hr test
and do not include the forced-impregnation gains or losses.

To determine the penetration of uranium and beryllium into these
rods, successive 1/32-in. layers were machined from the rods and sub-
mitted for chemical analysis. When the rods became too small to machine
without breaking, the ends were cut off and the center section was

submitted for analysis. Typical results are listed in Table 2.2.3.

89

Table 2.2.3. Analyses of Cuts Taken from Graphite Rods Impregnated
with LiF—MgF2 and Then Soaked in LiF-Ber-UFu

Type of Graphite
TSF AGOT ATL-32 '  ATT-82

e i Be T Be U Be U B
) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm)
1 3500 6400 2200 4300 1000 1800 600 1000
2 3300 6400 1500 3800 700 1200 150 270
L 3100 6300 1300 3300 550 900 25 85
8 3500 7000 940 2900  Ls0 600 10 100

Center 3700 5200 1200 1400 1900 450 9200 350

The cut numbers were assigned so that cut No. 1 was the first cut taken
on the rod, cut No. 10 was the last cut taken, and "center' indicates
the remaining center section, which was also submitted for analysis.

It appears that the ATL-82 and ATJ-82 grapites are resistant to
forced impregnation with LiF-MgF2 salt and are considerably resistant
to penetration by a typical reactor fuel. The most startling observa-
tion, however, is the unexpectedly high concentration of uranium in the
center section of the ATL-82 and ATJ-82 rods. Subsequent long-term
soaking tests of reactor-grade graphites TSF and AGOT in a fuel mixture
of the same IiF-BngeUFu composition also showed disproportionately
high urenium concentrations in the center in every case. Ifforts are
being made to determine the mechanism of this phenomenon and to further
verify it.

An experiment was carried out to determine whether untreated TSF
graphite could be forcibly impregnated with the LiF-BeFE-UFh (63-37-1
mole %) fuel mixture of current interest by using vacuum and pressure
techniques at the maximum reactor temperature of 125OOF. Six TSF
graphite rods were used in this experiment, and weight gains that
ranged from 0.1% to 1.85% were obtained. To determine depth of

penetration, these rods were machined, as previously described, and

90

the machine cuttings plus the center section were submitted for analysis.
In all cases, the bulk of the salt penetration was in the first and
second machine cuttings. HoweVer, it was noted again.that, in general,
the center sections contaified several times (occasionally 200 times)
more uranium than the last cutting taken at about 1/4 in. from the

center.

Radiastion Damage Studies

INOR-8 Thermal Convection Loop for Operation in the LITR
The in-pile thermal-convection loop for testing fused-salt fuel

in INOR-8 tubing in the LITR was operated in preliminary tests outside
the reactor, and satisfactory circulation of the salt (IiF-BeFE-UFu,
62-37-1 mole %) was obtained. Radiography of the fuel tube after these
tests demonstrated that no opaque material had deposited at the bottom,
as in a prev1ous loop test. Ll

Thermocouples have been installed on the fuel tube of this loop,
and the air annulus tube has been assembled around the loop. Detailed
examinations of the thermocouples were made, and complete records of
their condition, inecluding individual phbtomicrographs of each thermo-
couple, are available. This examination record will be used to correlate
thermocouple performance during operation in the reactor with the
initial conditions of the thermocouples. The information obtained will
be useful in the construction of Ffuture in-pile loops.

Modifications were made in the cooling-air control system at the
reactor to prbvide for indiVidual temperature control of various portions
of the loop. After resistance-heating elements have been installed and
the loop inserted in the outer can, the loop will be operafied in the
LITR.

In-Pile Static Corrosion.Tests

Two fuel-filled INOR-8.capsules, which were described previously,ll

were installed in the MIR and are being irradiated at a temperature of

MMSR quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 112.

91

1250°F. The fuel is LiF-BeF ,-UF), (62-37-1 mole %), and the power
density in the fuel is 1200 w/cm3. Two additional capsules of the

same type are being prepared for irradiation in the ORR.

Preparation of Purified Materials

Purification, Transfer, and Service Operations

The processing of small batches of various molten-fluoride-salt
compositions for use in corrosion tests, physical property studies,
and small-scale component testing decreased considerably during the
first half of this quarter. However, demands have gradually increased
during the past month, and a normal rate of operation is now being
maintained. A total of 35 kg of mixtures not containing beryllium
and 85 kg of beryllium-containing mixtures was processed during the
quarter. Transfer and service operations to assist engineering groups
in handling high-temperature fluids were continued at a slightly

increased rate.

Fuel Replenishment Tests

A simple device for testing the proposed fuel sampling and

enriching mechanism;g was constructed and tested. The device employs
two salt storage pots interconnected by a transfer line, with the
enriching device on one pot and the sampling device on the other pot.
The molten salt is moved alternately from one pot to the other with -
semiautomatic controls by using differential gas pressures.

- Solid, cast, UFM fuel slugs weighing 40 g each were introduced to
the dissolution pot in a copper basket and the salt, LiF-BeF, (50-50
mole %), was cycled between the two pots. The rate of dissolution was
measured by withdrawing and weighing the amount of UFh remaining;
samples of the solution were also taken for chemical analysis.

A 4O-g slug of UFh dissolved in 20 cycles (1 1/2 hr), but it was

not completely dissolved in 15 cycles (1 hr). This rate of solution

12MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268L, p Q.

92

appears to be adequate for convenient enrichment procedures.

Pure Compounds Prepared with Molten Ammonium Bifluoride

Molten emmonium bifluoride (mp, 125°C) has served as a reactant
for preparing both 51mple and complex fluorides. 13,1k Recent trials
show that commerc1al chromlc oxide is not noticeably attacked by molten
ammonium bifluoride. Magne51um.ox1de is converted into NHhMgF3, which
can be decomposed 1nto pure magne51um fluoride. Beryllium oxide
(calc1ned at 900 C) is also converted to an ammonium complex.

~ When commercial "Co 03” powder is treated for 1 hr with molten
ammonium bifluoride, a small amount of oxide is left unchanged. The
reactibn product has an x-ray diffraction pattern Somewhat shifted from
that of KCOF3, so it is presumed to be NHhCoF3

Nickel oxide was found to be more resistant to attack than ”00203
A l-hr treatment left almost all the oxide unreacted.

The oxides Ge02, UO,, ThO,, and CeO, seem to react completely with
molten ammonium bifluoride, but the products have not yet been identified.

Vanadium pentoxide reacts with molten ammonium bifluoride to evolve
yellow fumes. Presumably, Vanadium>pentafluoride is formed; it is
reported to react with moist air to form;yellow oxyfluorides.l5 |

For all the reactions with oxides, at least a 100% excess of
ammonium bifluoride was used in order to form a mixture which was
sufficiently fluid to stir. | |

Electrolytic iron powder reacted with molten ammonium bifluoride
to form ammonium hexafluoferrste (III). Chromium metal reacted to forfi
ammonium hexafluochromate (III), but the reaction was slow. Granulated
metal, finer than 100 mesh, was treated with the melt for 70 min, and

the unreacted ammonium bifluoride was volatilized. Extraction of this

138, J. Stuwrm end C. W. Sheridan, Preparation of Vanadium Tri-
fluoride by the Thermsl Decomposition of Ammonium Hexafluovanadate (III),

ORNL CF 56-5-95 (May 28, 1953).
My . Sturm, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 107.
0. Ruff and H. Lickfett, Ber. deut. chem. Ges. ki, 2539 (1911).

93

product with aqueous nitric acid to remove the ammonium hexafluochromate
(III) showed that 60% of the chromium metal had not reacted with the

ammonium bifluoride.

Reaction of Chromous Fluoride with Stannous Fluoride

The reaction of chromium metal with molten stannous fluoride pro-
vides a useful method of preparing pure chromous Jf‘luozc'n‘.d.e;llF however,
an excess of stannous fluoride can further oxidize the chromium. The
predominant phase resulting from the reaction of equal molar portions
of stannous fluoride and chromous fluoride is a compound thought to be
Cr3(CrF6)2.l6

The product obtained by fusing 1 mole of chromous fluoride with 2
moles of stannous fluoride appears to be stable at the boiling point of
stannous fluoride (~700°C). At 1000°C, all the stannous fluoride
volatilizes and leaves a residue of chromic fluoride. The probable
intermediate, Sn3(CrF6)2, has not been definitely identified, but the
reaction is presumably:

00°C

3
uSnF2 + 20rF2 ~—-———>Sn3(CrF6)2 + Sn

1000°C
SnS(CrF6)2 — 2CrF, + 35nF, 1

1°MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 113.

ok

2.3. TFUEL PROCESSING

Processing of molten-fluoride-salt reactor fuel by volatilization
of the uranium as UF6 appears to be feasible. The barren LiF-BeF2 can
be recovered for reuse by treatment with nearly anhydrous HF, in which
it is appreciably soluble, and in which rare earth neutron poisons and
most other polyvalent element fluorides are insoluble. Development work
on the process has continued with a study of the behavior of neptunium,
vhich is possibly the most important neutron poison other than the rare
earths.l

Further measurements of the solubility of Np(IV) have indicated
that its solubility (Table 2.3.1) in aqueous HF solutions saturated
with LiF-BeF-ThF) -UF) (61.5-37-1-0.5 mole %) will be considerably less
than reported previously. Apparently the Np(IV) solubility is decreased
by the presence of excess ThFu and UFM' The addition of nickel and
iron metals to solutions saturated with LiF, Ber, and Nthralso caused
a significant reduction in the neptunium solubility, except in anhydrous
HF. This is probably the result of reduction of Np(IV) to Np(III).

In a reactor processing system the solution will contain much larger
amounts of rare earths, plutonium, uranium, and thorium than of nep-
tunium, and the container will be a metal, probably a nickel alloy.
Therefore & neptunium solubility of the order of 0.0002 to 0.00005

mole % is expected in actual processing.

1MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 115.

95

96

Table 2.3.1 Solubility of Np(IV) in Aqueous HF

Np(IV) Solubility™

Np(IV) Solubility in HF Saturated

HE in HF Saturated . .
Concentration with LiF-BeFo with LiF-BeF,-ThF)-UF)
(%) mg/g of mole % relative Fe and Ni Added
solution to salt mg/g of mole % relative mg/g of mole % relative

solution to salt solution to salt

80 0.026 0.0031 0.0013 0.00012 0.0043 0.00052

90 0.011 0.0012 0.00046 0.00005 0.001k 0.00016

95 0.0086 0.00072 0.00054 0.00005 0.0029 0.00025

100 0.0029 0.00024 0.0047 0.00039 0.0029 0.00025

®Values previously reported in ref. 1 for LiF-BeF2 salt in absence of UFM and ThFu.
» * L L] * L

b E=EQsQrE-HdugEBHRD

gt
-y
- . * > L]

.

* * ] L ] *

GEPpUONGEIHOSTE S0 00

INTERNAL DISTRIBUTION

G. Affel
G. Alexander
S. Bettis
Billington
Blankenship
Blizard
Boch
Borkowski
Boudreau
Boyd
Bredig
Breeding
Briggs
Browning
Campbell
Carr
Cathers
Center (K-25)
Charpie
Coobs
Culler
. DeVan
. Douglas
Emlet (K-25)
Ergen
Estabrook
Ferguson
Fraas
A. Franco-Ferreira
H. Frye, Jr.
R. Gall
T. Gresky

L L *

fi?fi?dbmg"fflbbjfi.momsflQ?m‘#C{t‘*Uuflm

o

. L. Gregg

R. Grimes

Guth

S. Harrill

R. Hill

W. Hoffman
Hollaender

S. Householder
H. Jordan

W. Keilholtz
P. Keim

L6,
47,
L8,
4o,
50.
51.
52.
53.
54,
55.
56.

L] .

il 3 el = o JRSS

=]
.

c—({?jf;dtrfc—{

=00

G.

* L] - . L} - -

L 4 . * L -

'#U':UM‘PC-IOMP:E.’ME{QPMZQ*U

ORNL=-2723
Reactors — Power

TID-4500 (14th ed.)

:

Livingston
MacPherson
Manly

Mann

Mann
McDonald
McDuffie
McNally

Metz

Milford
Miller
Miller
Morgan
Murray (Y-12)
Nelson

Nessle

. Osborn
Patriarca

M. Perry
Phillips

M. Reyling

T. Roberts

T. Robinson

W. Savage
W. Savolainen
L. Scott
E. Seagren
D

J

* e e ] L]

» * *

L 2

munNzowa?wwpwumm>m

. Shipley
. Skinner
H. Snell
Storto
D. Susano
A, Swartout
Teboada
H. Taylor
E. Thomsa
B. Trauger
C. VonderLage

o7
89. G. M. Watson

90, A. M. Weinberg

91. M. E. Whatley

92. J. C. White

93. G. D. Whitman

ok, G. C. Williams :
g5. C. E. Winters

96. J. Zasler

97-100. ORNL - Y-12 Technical Library, Document Reference Section
101-140. Laboratory Records Department

141. Laboratory Records, ORNL R.C.
1so-1kly, Central Research Library

EXTERNAL DISTRIBUTION

145. D. H. Groelsema, AEC, Washington
146. Division of Research and Development, AEC, ORO
147-734. Given distribution as shown in TID-4500 (1kth ed.) under
Reactors-Power category (75 copies - OTS) |

98

Reports previously issued in this series are as follows:

ORNL-2373
ORNL~2431
ORNL~ 2L T4
ORNL-2551
ORNL-2626
ORNI-2684

Period Ending September 1, 1957
Period Ending October 31, 1957
Period Ending January 31, 1958
Period Ending June 30, 1958
Period Ending October 31, 1958
Period Ending January 31, 1959

99