ORNL-2474.txt

 

 

g

¥

¥

 

 

LY PROGRESS R

KL
"“-""

. BT -
SETLLLT :

 

 

 

 

.

Lok f

 

 
 

 

 

 

 

 

 

 

 

 

 

. tor' the Comm-ssmn, nor uny pflsm uchng on l:eholi oi the Commuslon.. .
A, Mckas nny wurrcnty or reprqsqmuflom ‘express -or ;mpltad with’ respecf fo !he occurucy,
,completoness, _or usefulness sf the in!ormflhan contamed in |h:s reporf or !hct fl:e use. of'

:"_emy mformut:on, upparoms, methcid ofprocess d:scloud ln this. report, - P
As -used in the ubove, person acflng on beho" of fhe Commlsslon mcludes nny ernployeg o.-.:‘." g
‘ comrucror of tho Commissnon to the -exte ;

 

 

 

 

 

 
 

 

S

)

 

ORNL.-2474 - _
UC-81 -~ Reactors—-Power

Contract No. W-7405-eng-26
MOLTEN-SALT REACTOR PROGRAM

QUARTERLY PROGRESS REPORT

For Period Ending January 31, 1958

H. G. MacPherson, Program Director

DATE ISSUED

lwlfiJ

MAY 141958

 

e omc RIDGE NATIONAL LABORATORY -

- Qak Ridge, Tennessee -
-operated by
UNION CARBIDE CORPORATION
for the

U.S. ATOMIC ENERGY COMMISSION

 
 

 

 

g 1ok e

"

 
.‘"*‘f

=) )

L2

o

FOREWORD

This quarterly progress report of the Molten-Salt Reactor Program records the technical
progress of the research at the Laboratory, under its Contract W-7405-eng-26, on power-
producing reactors fueled with circulating fused salts. The report is divided into two
major parts: 1. Reactor Design Studies and 2. Materials Studies.

Until July 1, 1957, the Molten-Salt Reactor Program was largely a design study, with
only token expenditures for experimental work, As of July 1, the program was expanded
to include experimental work on materials. A further augmentation of the program occurred
on October 1, 1957, when personnel and facilities for additional research and experi-
mentation became available. As a result of these transitions, this quarterly report has
been expanded to include component development and testing, engineering research,

metallurgical and chemical investigations, and radiation-damage testing.

 
 

 

 

v

 

 

"
a}

N

 

 

»

 

«)

3Y

 

CONTENTS

SUMMARY ettt e st s esa e e et sae sert se bR s abe Sa0m b sste s ot £ b en e s e e e RR e R bt e e s asbeEnR e arA s et e et sasaen s 1

PART 1. REACTOR DESIGN STUDIES

Tol. REACTOR DESIGN......ou oot ceererereseeeieretessssesteserenssresasesasessessasasbs sssnsesenssssasassssstas atanssssess sssassenssasanassssaeses 9
Study Layouts of Reactor System........ciriininicncinii it s issess e sassse s ennsnssnns 9

NUCIEAr CalCUIGHIONS ..ottt i s s s rsee et sese s e e esn e s asenases sunsosesbrnnnssnnass mbranaessessens 12
Univac-Ocusol Reactor Calculations ............ eeeeeestteasearaeat et bereaat e et ranses b e b et est s s et s sasae e s 12

Oracle Caloulations. ..t ieerarsreseere s eas s e ssasssteeseemesrse s bessas e sbassssassesarsssnensassees 14

IBMo704 COURS v.veureiireinenreiencreesiereseuree et rerrresssesaesaass seestssasasesss sasasssssassessstonsstaserasstessare ssnenessnrsens 15

Fuel Fill-and-Drain System Design ......cc.vieeeiniiintiiintnisstssssese s ssasss s sesssssasssssnns 15

Design of a Hydrostatic Bearing for a Molten Salt Pump Impeller ..., e 16

1.2, COMPONENT DEVELOPMENT AND TESTING.....cceoriinerrneeneceecrennameeenessssea st scnssss s sasnssnsses 18
Fuel Pump Design and Development .........cccuiiiiiinniiniiiciiinn s svsnsaenssiessssssnsesanseasas 18

Pump Design Studies ... sttt snas st s e e san s e n s e nas 18
Development Tests of Salt-Lubricated Bearings......cccoviviiminiiiiininiiniiessnens 18
Development Tests of Mechanical Shaft Seals ..o 19
Development Tests of Oil-Lubricated Bearings ......c.cccovvrvvnrrercrrnccnnncn. rereerereeteteiaate e sseenans 20

Irradiation and Endurance Tests of Bearings and Seals ....uooeveivcerrniiniiiincinncicencnn, 20
Development of Techniques for Remote Maintenance of the Reactor System ......ccevvnrcenrnencnes 20
Mechanical Joint Development ... ... eriicrinec e e esenae e tisenessssns s s ssessnsnns 20

Remote Manipulation Techniques .....cciericiicncees vt e s b be s s a s aees 24
Heater-Insulation Units for Remote Application ...ccccevirecirceneriiniiirnen e 25
Maintenance Demonstration Facility.....oeeivrrnereinrre ittt st st sre s eessssrsenes 26

Heat Exchanger Development ..........coccmiiiiviniicii st s s st rasss s st s ssesssssenans 27

Design, Construction, and Operation of Materials Testing Loops e reeuseesenasssbesrase s srassnesastsans 27
FOrcedeCirculaHOn LOOPS oeuvrivrrireiteece e srttteesesssiresanearessssesaresasesaes sesssssnmsssessensssssrasassssnsnans 27

in-Pile LLoops .coccvecccirecennens Ceuerusasheneterenaa e e e e R E B SR e sS4 SO A SR R SRR SRR e 08 32

1.3. ENGINEERING RESEARCH -ovevrssecssseserrensern e gen e s 35
Hydrodynamic Studies of MSR Core e rienreneerasase rreseeeserte e irsnerbe e reneher et e b npanesa et se s area sas veeersneas 35
Physical Property Measurements ......ccivoceenivescssencees renenatasainaransnesassansrenin \evesieasasaeiestesarensizasasensns 37

Molten Salt Heut Transfer Studie's torsosentsivatrerebaisRR L ERRs P AL R 0RO 4 RS S8 SR e SRR b 088 ES 37

1.4. ADVANCED REACTOR STUDIES 1vvcerermrserrnes et e et s s eenee 41
A Molten Salt Naturcl—Convechon Reactor ..... eieerieenararessereas eseesesere s rene e et st e st e e ra s be e nanenssaass 41
 Salt-Cooled Heat Exchangers .......coviemsrumimuesnssivnnnsnisssonna: eestereancpeses et e rsa ks peR bbb r R et erere st eaas 42

. Helium-Cooled Heat Exchangers for Steam Cycle ..................... 42
Helium-Cooled Heat Exchanger for Gas-Turblne Cycle ................. treresssisvensfosepissnaranssomssasst s 43
Companson of the Various Coohng Sysfems ..... Cisvinevessnsisssasass eterestsessseesseseneesansesrreonrassntoaran, 44

' Compurlson of Nafura|-Convecflon System with Forced-Convechon SyStem..umvereeriarerrraeens 44

ngh Flux Reacfors .............................................................................................................................. 44

 
 

 

 

~ PART 2. MATERIALS STUDIES

2.1 METALLURGY ocerenoerms oottt s sttt s oo 51
B Dynamic Corrosion Studies .......... ................... 51
Thorium-Bearing Salts in Thermal-Convection Loops .....cccccecevireucccrerivenerenens enenerases s annans 51
Uranium-Bearing Salts and Coolant Salts in Thermal-Convection Loops: .....eceeurrvarsrermenenne 52
Results of Examination of Samples Removed from Forced-Clrculaflon Loop CPR ............ 54
General CorroSion STUIes.... ..o vieiiiinnctcse et ce s seeerens setens e st s e asresstsssnsasesnsnsssssnssnasasssasans 54
Effect of Carburization on Reactor Structural Materials .......oooverviiiiiinniniiincccccninee 34
Corrosion of Brazing Alloys by Fuel Salts .ot erere et seenee et aes 59
Corrosion of INOR-8 Welds by NaK and by Fuel Salts....ivriiiisieniriicnnsinnns. 60
Physu:cll Properties of INOR-B ........... 62
Mechanical Properties of INOR=8.......c..coccuiriereierivscsissnssssesmssessessssssssessimssrsssssssessasesssses sesnssssorssn 63
Welding and Brazing Studies.....c..cccviverrnnnes vemesiestsEeneneesnete stssneneiote sermsnneca sbe s bsmnn esisatrest siem e s as st asen 64
Metal Seals for Remote-Disconnect Flanged Jomts ................................................................ 64
Welding of INOR=8 TUDBING ..cccoueveminrcerncieinneestininssccresccsasnest s ssssnsnsessesessssssnsssssssesssssosanssass e 65
Evaluation Tests for Welds ...coeveeverereccnrnnene. ceeaeeeersta et et srennnsenans eereenesaerseesesenransanssasrees 67
Examination of INOR-8 Forced-Clrculoflon Loop That Failed During _
NGl HEGHRG coveveereeriieceecrereneeee e senscesaseresnsnssssness s stsssssssse ssrsasssesssssssnsnsesnsnssessnsans oos SR 71
Fabrication of Test Components ... cienieneenecsenrereseiesieresnsssessesses ssessesassssssons vereseneensneen 73
Development of Nondestructive Testing Techniques ............... eteresener et et s st et sa e e R n s aerasaensanane 75
Evaluation of INOR=8 TubiNgG....ccccceiieerierecrereriimneierssniameesesaessnsssnsssessnsssnsesnessessnsssasssessenssssesses oe 75
Cladding Thickness Measurements and Bond |nspechons ...................................................... 77
INSPECHION RESUIES oottt et e sanens et s e e s assessens oo snasnasennesnssbanesbnsnesesstanses 77
Material INSPECHION ..t re st eesae s e res srasae e basssesaesees e gesmssassssortssresnanss ebesnss 77
Weld Inspection ....cccvvenvvivvnivreesinnenna. fevereestnteterseesaneratetesreeaseeaeanteertsashnetarnaebastea s tesasesnanaernnteanens 78
Failure AnGlYSes ...ccooueiiiieieiiicieciteceescseree st eiae e sebestsnesssesesssesssasses s ssnn st bssassarasmesssnsnensansinssnnnss 78
2.2, RADIATION DAMAGE .....ooceireecreeetetisreres st setestetesss et eassssss et evesbasbeserssssasssssesens evnereree st et rae s 79
In-Pile Thermal-Convection LLoop Tests ... cnseeseerrsnesessssissesssarsesssessesnsresssseseasens 79
INcPile Static Capsule Tests ettt reveessrsessesstsst ssssessesersasnenpessssassrnssessons 80
2.3, CHEMISTRY wooroesoeseevssrssssmssmssesssssmssssmssssssssssnessesssssss s e s s 81
Phase EQuilibrium StUdies ...t s riaee s siasssssstsssssstssenssssnssssssesasssssrnsnens 81
Systems Containing UF, and/or ThF ; oottt 81
Solubility and Stability of PuF, in Molten Flyerides.......... etereearesteneaestanteasereesavares antasesnsassen 88
Fused Chlorides as Secondary Heat Transfer Fluids c.cuviveciniinniiniccsniitcic s 90
Yapor Pressures of LiF-BeF, Mlxtures .......................................................................................... N
Fuel Reprocessing....cccomerpeinennininisnennnsicncsianeesssosnnssessesersessssesssass veranresaetaesasreteseresasnssnanes 91
Solubility of Noble Gases in Molten Fluoride MixtUres .......c.veeeemeeemreereesesesssscersessssereensens 91
Solubility of HF in Molten Fluorides ......rencniescncsriecnnenees SO 93
Solubilities of Fission-Product Fluorides ... 94
- Order of Oxide Precipitation in Fluoride Salt Melts......cccocevvverreccriveccrnnne reeveenereeeesnraseres 97
Chemical Reactions of Oxides with Fluorides in LiFeKF .. - 99
Lithium Recovery from NaF-KF-LiF Melts .......ccocovirommreomnerrreessosessmsosessssssenesssssssseees cereeras 101
'Vvi

13 )

3 )

%)

 
»

«)

#

 

 

»

*)

 

 

Chemistry of the CorroSion Process .........iccriieecseeriesessessisssssmessssssisessssssessssssssssssssssassenes 105
Activity Coefficients of CrF, in NoF-ZrF......cooeece... reeierenesrater et enensaereans peesesresenesesanenses 105
Solubility of FeF, in LiFeBeF 5 ceeecs st sttt arnes 107
Use of Cr3! to Study Chromium Migration in Polythermal Inconel—

 MOoHEn Salt SYSIEMS .o.ceiveirice st s ras s srsasr et e e s easn s s sarens 107
Activity of Nickel in Nickel-Molybdenum Alloys .....ocoirvrerrcineeerecene s m

Production of Purified Mixtures ............. reeserenerresteaesrnerasnsenns rereneeenne rerereseessserserrresesiaseenessesnsnnesassane 113
Preparation of Pure Fluorides ..........coiiiicvrivrrinennneirecnemnnsessnsssine s sseesssssssssssssessnens 113
Small-Scale Purification Operations ........ccoeeuverieemssiseeennn s snssesuesnsaresaserenseanssesesensrasennsasrenss 113
Preparation of Material for InaPile Loop ... cseesesesae s ieesasne s 113
Tronsfer and Service Operations ........cucueene. rieaeseeeseiearseseeessestresseestesanasbessseabeshesentennrtstennrennean 113

vii

 

 
 

 

J

*)

NIRRT RTE S

)

 
"

«)

N

Y

N

)

 

'be possible to exceed 1.0,

MOLTEN-SALT REACTOR PROGRAH QUARTERLY PROGRESS REPORT

SUMMARY

PART 1. REACTOR DESIGN STUDIES
1.1. Reactor Design

Preliminary layouts have been prepared of the
molten-salt reactor system that are based on the
use of five fuel pumps and two blanket-salt pumps,
The total thermal power production is assumed to
be 638 Mw, with 90% of the power being generated

‘and leakage probabilities. The resulting constants

reproduce the initial spectrum ond critical con- =

centration exactly, ' S
A preliminary fuel flll-cmd-drc:m system desngnr'

- 'was completed which satisfies the design criteria -

in that (1) it is always in a standby condition in

which it is lmmedlately available for drainage of
~ the fuel, (2) it can adequately handle the fuel

in the 8-ft-dia core and 10% being generated in the |

2-ft-thick blanket. The fpfopos_ed layouts are being

afterheat, and (3) criticality cannot be achieved in

- the drain system. The fill-and-drain vessel con-

studied in order to achieve simplification and to

minimize the fuel inventory,
reactor cell are also being studied in order to
determine the best possible arrangements of piping
and other components that . will provide minimum
plant size and yet facilitate remote maintenance..

The parameter study of two-region, homogeneous,
molten salt reactors was continued through nuclear
calculations on the Univac, the Oracle, and the
IBM-704, The Univac calculations indicate that,
for the same core diameter and thorium content of
the fuel salt, the U233 inventory would be one-

half the U?35 inventory., In the U235 case the
regeneration ratio is limited to a moximum -of

Layouts of the

sists of forty-eight 20-ft lengths of 12-in.-dia pipe :

arranged in six vertical banks connected on -

opposite ends with mitered joints, The system is
preheated and maintained at the desired tempera-
ture with bayonet-type electrical heaters in tubes
located axially in the pipes. Removal of afterheat
will be accomplished by radiant heat transfer to
‘banks of wuter-f:lled boiler tubes that will normally
be dry.

A hydrostahc bearlng was desugned for use in

fuel pumps which differs from the conventional

bearing in that the pockets rotate on the impeller.

- A bearing of this type has the advantage that the

about 0.675, and in the U233 case it appears to |

It is pointed out,

however, that the regeneration ratio for a y23s.

fueled reactor will decrease with time as burnup

poisons accumulate, while in the U235 fuelec!_,_'
reactors with initial regeneration rahos ‘of -about .
0.6 or better the regenerqhon ratio may actially "
lmprove with time: and with reasonable chem|c0|-?-‘. .
processmg rufes because of the bu:ldup of the_li_f-

superror U233 fyel.

Comparlsons of - hth:um-bery”lum ond sodlum-;:r_‘i..r'i: E
beryllium salts showed that the’ sod:um-berylhumf';'_; o

salts’ requtred 1.1 10 2 times the U2 ¥

for computing the constants from the concentra-
tions, absorption cross sections, initial spectrum,

inventory .~

"ireql.nred with the lithium: berylhum salts, Atso,""';-ffi_;'fi
: V,ffhe sod;um-berylllum salts show a dlsadvanfage off '_, S

0.11t0 0.15in regeneruhon rcmo. S T

Mod:ficatlons were made in the borghum program--
'3__for the Oracle, which” was: desrgned for computing. -
progressive changes in “core . concentration “and. .
regeneration ratios. A subroutine was mcorporated“‘ '

 

,,?_'.;';-._,::_Pump shah speed

 
  
 

pressure of the pumped fluid would maintain the |

_centering of the impeller in the pump casing. B

12 Component Developmenf and Tesfmg

Prehmmcry pump design studies have mdicated

- that five pumps with the followmg characteristics

- would be’ suitable. for cnrculchng fuel 130 (BeF -

- LiF-UF , 37-62-1 ‘mole %) in the moifen salt
reactor bemg consndered Lo -
V'Z-tFlow through pump 7 o '_'_":4590 QP"‘
,;,;Head produ::ecl by pufip ‘ 7'! ft
: Ei;ig‘;_Temperature uf pump mlet -':"'1230°F (max) _
| -Sfohc pressure af pump mlet 19. 2 psuu o o
o 7..7']160 rpm e

 

f"Pump desngns and arrongements ‘are bemg studled o

‘ ir':“ order to. determme the most favorcble arrange- - -
ment ‘of cH elements w:th respect o adapiablllty e
“to reactor consfruchon, durability, remote mainte-

nance problems, reactor operational probiems, and
other considerations,

 
 

 

 

 

 

e et e e e mein

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Facilities are being designed and constructed

for development tests of the salt-lubricated

bearings being considered for use in fused salt

pumps. The oil-lubricated bearings being used at

present are unsatisfactory because they must be
shielded from both the heat and the radiation from

~the fuel salt in order to avoid damage to the

bearings and to the lubricant, Equipment is to be

set up for -high-temperature tests of both hydro-

dynamic and hydrostatic salt-lubricated bearings.
Oil-lubricated mechanical shaoft seals are being

_ studied for auxiliary system applications. Modi-

fications in purge-gas fiow are being tested as a

means of preventing diffusion of process-fluid
~ vapor into the seal region, Also, since oil-

lubricated bearings may be required for the upper
seals of the fuel pumps, tests of Dowtherm A as
a lubricant are under way, This lubricant decom-
poses  into gaseous and liquid products when
irradiated, rather than into carbonaceous deposits
of the type that result from the irradiation of
hydrocarbon-base lubricants, A fest of a pump
rotary element with oil-lubricated bearings is also
under way in a radiation field. Further, a bellows-

mounted seal is being operated in an endurance

test in a sump pump that is circulating NaK at
1250°F,

" Techniques for remote maintenance of the com-
ponents within the reactor cell are being studied.

1t is considered that the most feasible solution to

the maintenance problem is to make all components
removable and replaceable by remote manipulation,
Therefore means for remotely separating and
joining pipes are being investigated. "Tests are
under way of three types of flange joints: (1) a
freeze-flange joint in which a frozen seal of molten
salt is used to prevent leakage of molten salt
thraugh the space between the flanges, (2) a joint
which has a cast-metal seal between the flanges,
and (3) o joint in which a V-shaped tooth on each

: flange indents the sealingring.

The various types of remote-handling appurafus

‘now ‘available commercially are being studied for

applicability to maintenance of this reactor sys-

tem, - The remote assembly and disassembly of a

pump in a hot cell is being attempted as a means

 of studying the manipulation problems. Design

work has also been initiated on heater-insulation

“units that can be remotely removed aond replaced.

Plans are being made for a large-scale demon-
stration facility in which fo test mechanical joints,

the replaceability of components, the adequacy of
heater-insulation units, the unitization of wiring
harness and service piping, and the application of
remote-wewmg ond -hundllng apparatus ond tech-
niques,

A heat exchanger test facnllty is - bemg operated
to obtain heat transfer correlations for predicting
the heat transfer performance of the molten salis
of interest, Experimental information is required
because the heat transfer characteristics of some
salts appear to be affected by the type of struc-
tural material used and by the wettmg properhes
of the salts,

" Forced-circulation Joops in which large tem-

~ perature drops can be achieved are being operated
to study the corrosion of INOR-8 and of Inconel by
the fused salts of interest, The long-time effects

are of particular interest, and satisfactory opera-
tion of a loop will imply operation for one year or

- longer without significant equipment difficulties

or changes in operating conditions. Two special
loops have been constructed of INOR-8. One of
these includes graphite specimens and will be

examined after operation to determine the extent

to which graphite causes carburization of INOR-8

and the effects of the fused salt mixture on’

graphite, The other loop includes INOR-8 speci-
mens for weight-loss studies.

A forced-circulation loop is also being assembled

that will be operated in a beam hole in the MTR,
Operation of the loop will provide information on
fuel stability and corrosion of INOR-8 under
irradiation at simulated reactor conditions.

1.3, Engineering Research

Three entrance-exit systems proposed for the
core of the molten salt reactor are being studied
in glass models., Preliminary qualitative data on

flow patterns and velocity profiles have been ob-
tained through visual observation and photographic

recording of the motion of phosphorescent particles.
The three systems being studied consist of (1) a
core with the inlet and outlet diametrically oppo-
site each other (straight-through flow), (2) a core
with the inlet ond outlet concentric and the fluid
entering through the inner pipe and exiting through
the outer annulus, and (3) o core with the inlet

—and outlet concentric and the fluid entering through

the annulus and exiting through the inner pipe.

In initial experiments with the concentric system

and the fluid entering through the inner pipe, @

¥

¥

 
Hhme

(M

 

 

o)

AL e st b v o e s

.fli

C

primary heat exchanger,

 

 

large toroidal eddy was noted in the region be-
tween the high-velocity, central, downward jet
and the return stream along the core walis. Such
an eddy could engender discrete high-temperature
masses within the main fluid flow ond thus give
rise to high-frequency thermal cycling of system
components. The effect of o vortex generator at
the entrance will be studied. '
Experimental determinations are being made of
the viscosities and thermal conductivities of
several BeF ,-bearing fluoride salt mixtures. Heat
transfer studies are being made in order to de-
termine the effect of nonwetting and interfacial
film formation on heat transfer in Inconel and
INOR-8 systems.,
used for these studies with the salt m|xture

LiF-BeF,-UF , (53-46-1 mole %).
1.4. Advunced Reactor Studies

A  preliminary design study was made of o
natural-convection molten salt reactor which could
be used in a system in which there was a premium
placed on reliability and ease of maintenance.

The advantages of eliminating the fuel-circulating

pumps and the attendant problems of maintenance
are obtained at the cost of the increased fuel
volume required for a system in which the pressure
losses must ke very low. A gas-turbine cycle or
one of several steam cycles that operate efficiently
under high-temperature conditions would be used
with this system. In this study both helivm and
a molten salt were considered as coolants for the
A comparison of the
natural-convection system with the forced-circula-

tion system indicates that, for o 60-Mw . (fhermal)3"'j_"
reactor,: the naturcl-conVecflon sysfem requires o =~
fuel inventory about 42% grea’rer than fhat of the*___

forced-cuculchon system.

An idealized reactor ‘model -is bemg used in ak,i_»fi‘;
study of the influence of vcnous_' -

sys tematic

- factors on' the power required to obtain a given -
Data have been ob- -
_ tained -for the ideuhzed model, and - further studlesr"j'
- “are under way for'a model modlfled to reflect more -
'frealisnc condmons. - - e

flux in a resecrch reactor,

PART 2,, MATERIALS STUDIES
s - 2 ‘I. Meiu”urgy ' .
Cofrd'sion
thermal-convection and forced-circulation loops
fabricated of INOR-8 and Inconel.

of a three-phase test program has been almost

- reactor structural materials.

A single circular tube will be -

_ embrmle

the strength of INOR-8.-
':Vperlods are requnred to obtmn dofa on plashc
_:‘_,propertles, prellmmary data on fensule properties
- are being obtained for use in desugn studies.
'Duta are presenfed for sheet spec:mens that must-.
_  'be considered as _approximate. -
expenments are under way w:fh"

The first phase

PERIOD ENDING JANUARY 31, 1958

completed and part of the second phase of testing

is under way, Only Inconel thermal-convection
loops have been examined thus far, but the low
corrosion rates expected in these 1000-hr tests
were found, ,
Studies are under way for determining the effect
of carburization on the mechanical properties of
The sodium-graphite
system was found to be a rapid and effective
medium for carburizing stainless steels, Hastelloy
B, and Inconel. On the other hand, Inconel exposed

to the graphite—fuel 30 (NaF-ZrF -UF,, 50-46-4

“mole %) system for 100 hr at 1500 and at 1250°F

in seesaw-furnace apparatus did not become
carburized.  Static capsule tests revealed that
INOR-8 carburized more readily in sedium than
Inconel, The 'INOR-8 tensile specimens to be
used for strength and elongation determinations
are being prepared by using the sodium-graphite

.system,

- The precious-metal brazing alloys, 82% Au-18%
Ni and 80% Au-20% Cu, being considered for use
in the fabrication of fuel-salt—to—coolant-salt heat
exchangers were corrosion tested in fuel mixtures.
Neither alloy showed attack after 2000 hr at
1200°F in stotic .tests and 500 hr in seesaw-
furnace apparatus, Similarly, no attack was found -

“on any of the welded INOR-8 plates tested in NaK

and in fuel 130 in seesaw-furnace apparatus for

500 he at 1200°F. Various mckel-molybdenum-

base welding rods were used.,
Measurements were made of the modulus™ of
elasticity, the thermal conductivity, and the

tensile properties of several commercml air-melted
“heats
- determining whefher INOR-8 has « tendency fto
. in the temperature range of 1000 to -~
- 1400°F. Specimens are being aged for penods of
500, 1000 12,000, 5,000, and 10,000 hr, Pre-

||mmury results show that the specumens aged

‘of INOR-8,

~Studies were initiated for

500 hr did not become brittle.- o
Tests are under way for obtaining basic data on .
~'Since relatively fong

Relaxaflon fests
bemg ‘made to determine whether INOR-8 wili
deform plashcally under reactor operating con-
ditions have indicated that the creep strength
will have to be investigated.

 
 

 

 

- different heats of INOR-8 from three different

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

In support of the component development program, presence of the Li® isotope in the fuel mixture
metal seals for remote-disconnect flanged joints.  (fuel 130) for the in-pile loop. Since the possible
were investigated. A series of tests showed that effects of traces of tritium in the loop could not be
silver or a_ silver-copper eutectic alloy would evaluated with confidence, it was decided to use
make effective cast metal seals on flange joints the best available Li.
made of stainless steel, INOR-8, and Inconel. Preparations are being made for the opercmon of

Investigations of the welding characteristics of a similar loop in the ORR and for the irradiation of

INOR-8 tubing were continued, and evaluation fuel 130 in INOR 8 capsules in the MTR
tests were made of various weld metals. Seven

sources have been studied, and only one has 2.3. ChemIStry
shown cracking tendencies when welded. In these
investigations, weld test plates are prepared that
provide specimens for mechanical property studies
of welded joints, for radiographic, metallographic,
and hardness studies, and for obtaining general
information on the welding characteristics of the
materials under conditions of high restraint,

Phase equilibrium studies are being made for
determining whether an'l'.iF'-BeFi, mixture will
dissolve sufficient ThF, and UF, to provide a
fuel for a fused salt breeder reactor, The studies
have indicated that the quaternary system LiF-
BeF,-ThF,-UF, can be tredted as a temary
sysfem and that some interpolations can ke made

An examingtion was made of the INOR-8 forced- between the systems LiF-BeF,-UF, and LiF-
circulation loop that failed during initial heating Ber'ThF.q with regard to liquidus temperatures
in a test stand, The failure occurred near the and phase relationships, Breeder reactor blanket
fusion line of the weld joining o Hastelloy B o breeder reactor fuel solvent compositions,
nipple and an INOR-8 (Haynes heat SP-16) adapter. whose maximum ThF, concentration is limited to
The lack of ductility of the Hastelloy B nipple, that available in salts having less than a 550°C
which was part of a finish-machined Hastelloy B liguidus, may be chosen from an area of the phase
pump barrel, was the cause of the failure, INOR-8 diagram in which the upper limits of ThF, con-
pumps are now being fabricated for use in the centration are obtained in the following compo-
forced-circulation loops. sitions: 75 mole % LiF~16 mole % ThF ,~9 mole %

BeF,, 69.5 mole % LiF-21 mole % ThF ,-9.5

2.2. Radiation Damage mole % BeF,, 68 mole % LiF-22 mole % ThF ,~10

Preparations are being made for the operation of mole % BeF
an INOR-8 thermal-convection loop in the LITR. The 501U5||'*Y and stability of PUF in beryllium-
An electrically heated full-scale mockup has been containing fluoride saits are being mveshgafed
assembled and filled with fuel to test components The solubility has been shown to increase with
and procedures, A new type of thermocouple increasing BeF, concentration in LiF-BeF, mix-
assembly has been designed for use with this tures. In the NaF-BeF, system, the solubility of
loop. In mockup tests the thermocouple has PuF; is higher in the mixtures with 50 mole %
survived dozens of thermal cycles in which the BeF, and 36 mole % BeF, than in the mixture
thermol expansion was many times greater than with 43 mole % BeF,. Thus there is an indication
that expected in the loop. The new assembly that the solubility of PuF, in this solvent goes
allows the thermocouple jacket to move as the through a minimum in the vicinity of mixtures with
fuel tube moves during expansion. 43 mole % BeF,. The solubility .of PuF, ot
Charcoal for use in a trap for adsorbing xenon 565°C in the binary mixtures studied varied from

during operation of the loop was baked in vacuum  about 0.2 mole % for the 57 mole % NaF—43 mole %

for 48 hr at 500°C in order to decompose organic BeF, mixture to 0.45 mole % for the 51.6 mole %

impurities and was then fested with radiokrypton L|F-48 4 mole % BeF. mixture, This concentra-
to determine the effect of the heat treatment on tion range would probably be adequate to fuel a
its adsorptive qualities, It was found that the  molten salt plutonium-burner reactor,  in the one

_ charcoal was 10% more effective than it was ternary mixture studied, NaF-LiF-BeF, (56-12-28
before the heat treatment, mole %), a value of 1.5 mole % PuF, was obtained
Calculations were made of the magmfude of the at 565°C. Thus there is an indication that the
undesirable effects that would result from the solubility of PuF, continues to increase with

e

%)

 
 

"

AN

 

M

an

 

 

 
 

sy

decreasing BeF, concentration. No evidence of
disproportionation of PuF, has been found in
these experiments.

survey was made of the physu:al chemical,
and nuclear properties of fused chlorides of
possible interest as secondary heat transfer
fluids, The survey showed the eutectic composi-

tion 41.7 mole % RbC|-58.3 mole % LiCl to be

the most attractive from the standpoints of vapor
pressure and corrosion. Thermal-convection loop
tests would be required to determine the rate of
mass transfer. -Apparatus is being constructed

for treating RbCI-LiCl mixtures to remove the

water that is always present in the salts,

The vapor pressures of LiF-BeF, mixtures are
expected to be low at MSR temperatures, and, to
determine the magnitude, measurements were
made of the vapor pressure of the solvent mixture
64,9 mole % LiF-35,1 mole % BeF,. Since be-
havior similar to that of the NaF-BeF sysfem can
be expected in the LiF-BeF, system, it is antici-
pated that the vapor pressure of a 70 mole %
LiF-30 mole % BeF. solution will be about one-

half the vapor pressure of the 64.9 mole % LiF-

35.1 mole % BeF, solution at the same tempera-
ture,

Studies of the solubilities of the noble gases in
NaF-KF-LiF mixtures were continved. The
studies of the NaF-KF-LiF mixtures were under-

taken pending completion of facilities for studying

solvents containing BeF,. It is expected that the
numerical values of the noble gas solubilities in
solvents containing BeF2 will be less than the
corresponding values in NaF ZrF, ond “more. fhan

“ those in NaF- KF-LlF

Studies of the solublhty of HF in NoF-ZrF and

~ NaF- KF-LiF mixtures were continued, and studles‘_f
- of L:F-BeF2 mixtures -were - mmated " Both the":-ii
HF solubility and the heat: of soluhon vulues for -
. the LiF-BeF, (51-49 mole %) mixture are - of the -
. “same order of magmtude as the values obtamed .

7'__Wlfl'l correspondmg NaF- Z.rF4 mixtures,
_ The solubilities. of - flssmn-producf fluorldes ino-
- BeF -containing solvents ‘are being studied by
% ;)UF ‘systems is descrlbed The results of actual
‘ ’iloop expenments in which’ Cr3! was used to trace
© “the “chromium " migration were “used to check the
f}‘valldlty of the calaulations.

"usmg radioactive fracer. techmques. “Values™ are -
7"presented for the so?ubn!mes as.a funchon of'-‘.’;_f.‘i-
- '_"temperature and’ composmon of CeF3 in NQ::F-BeF2
~ YF; in NaF- BeF Cer in LiF-BeF; and CeF, -
‘ and LaF in 1he presence of each other in LiF-"“' =
of nickel in nickel-molybdenum alloys by using an

BeF
The precipitation of oxides of flsswn products
from fluoride melts is being studied as one of the

PERIOD ENDING JANUARY 31, 1958

possible methods for the purification of fuel
mixtures.  Simple thermodynamic considerations
are being applied in order to calculate the relative
order of precipitation of oxides. Calculations are

descrubed for the oxide precipitation of U**, Ce***,

-~ and Be*t from an LiF-BeF., mixture containing

UF, and CeF,. As part of this study the solu-
bilities and precipitation reactions of a variety of
solutes are being studied in the simple binary

solvent LiF-KF. The following usefu! generali-

zations are based on the information obtained thus
far regarding the solubilities of oxides of uranium,
zirconium, hafnium, rare earths, alkaline earths,
and alkali metals in this solvent,

1. Urenium, zirconium, and hafnium precipitate

as the dioxides and are the {east soluble.

2. Rare earths precipitate as R’2O3 and have

- very low solubilities.

- 3. Beryllium and magnesium oxides are slightly

“more soluble than the rare earth oxides but are

still quite insoluble, ‘
4, Barium, strontium, calcium, potassium,
sodium, and lithium oxides are more soluble than
the oxides specified in items 1, 2, and 3.

An experimental study of the activity coeffi-
cients of CrF, in NaF-ZrF , (353-47 mole %) was

continued, and equilibrium quohents obtained at

~ 850°C for the reaction

CrF,(d) + H,(g) == Cr(s) + 2HF(g)

are presented. Experlments are under way at

‘other temperatures.

~In.. preparation- for a determination of activity

coetf:cuents ‘of Fer in LlF-BeF {63-37 mole %),
it was established that ethbrlum measurements
»'cun be made with solutions of FeF2 in this solvent
“af concentrahons ‘well in excess of 5000 ppm
“Fe**, even at 500°C which is the iowesi tempera-
ture of interest.

- A proposed graphlcal method evolved for calcu-
Iatmg ‘the rate (_:md amount of chromium migration
(corrosnon) to be expecfed in Inconel ~NoF-ZrF -

Afiempts are -being made to measure the actw:ty

electromotive-force method, ~ Experimental diffi-
culties are being studied.

 
 

 

'

 

e

T

 
1 -

3

b

"

 

B

"}

 

 

Part 1

REACTOR DESIGN STUDIES

 
 

 

A

 

o

 

 
€y

¥y

+)

 

 

REACTOR DESIGN
H G MacPherson

STUDY LAYOUTS OF REACTOR SYSTEM
E. J. Breeding |
J. Y. Estabrook -~ W. S. Hcms R. E. Helms

Prellmmary layouts of the molten-salt reactor
(MSR) system are being prepared as a means of -

studying the mechanicel design. These layouts
are based on the use of five fuel pumps and two

blanket-salt pumps.  The total thermal power.

production is assumed to be 638 Mw, with 90% of
the power being genergted in the 8-ft-dia core and
10% being generated in the 2-ft-thick blanket,
Each of the layouts, presented here as Figs. 1.1.1
through 1.1.5, has both satisfactory and unsatis-
factory features. - It is hoped that through examining
many possible arrangements the most simple de-

sign can be determined. Layouts of the reactor
cell are also being made in order to determine the

best possible arrangements of piping and other
components that will provide minimum plant size
and yet facilitate remote maintenance, |

The basic layout shown in Fig. 1.1.1 of a two-

pass reactor with a blanket was conceived with the

support and fabrication problems as the major
influence. The main support ring is, in effect,
a channel rolled into @ ring, Both the blanket and
core shells are supported by this ring. The fuel
pumps are supported by their barrels, which pene-
trate the top flange. Both the fuel and the blanket-
salt expansion tanks are formed by conventional
shell-head shapes. Individual pump suct:on hnes

direct to the neck of the reacfor core vessel are
prowded to avoid stresses ‘induced by penetrchon.
of the biankei shell The ‘method of support of -
the pumps - permits . the pump discharges to be -
rotated without altering the details of. fabrlcuhon, o
of the vessel, Except for the outrigged. blanket-_ e
salt pumps,_the unit is symmetrical in order to - -
-_\_s|mpltfy fabrication. All parts ‘of the fuel systemi U

" with blanket salt, and yet simple shapes have been

retained for ease of fabrication. The fuel pumps
are  equally spaced around ‘the periphery of the
reactor, -and the two blonket-salt pumps are located
in a central annulus ‘of blanket over the core.
This arrangement permits the support of the pumps
gnd the blanket pressure shell from a grid type of

UNCLASSIFIED
ORNL—LR—DWG 27883

 

    
  

FUEL BLANKET-SALT
PUMP EXPANSION TANK

   

 

 

 

~BLANKET-SALT
~ OUTLET

 

 

 

are in packages of five in order to lower the cost =

-of fabrication. The fuel lnventory of .a reactor of LY
~this ‘type would _be low, that is, ‘approximately = .-
330 ft3 One of lhe main’ ob|echons to this iayoutf S
“is that there isno blanket sclt around the neck of .~ l

the core vessel,

The layout presented in Flg. 1.1.2 has the . -
advantage that the core is completely surrounded -

   

~-BLANKET-SALT INLET

Fige" _'l.i.-l. MSR Layout with Outrigged Blanket-Salt
Pumps ond No_'Bl_t:fikei Salt Around Neck of Core.

 
 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED
ORNL-LR-DWG 27884

 

 

 

 

 

 

 

FUEL —
QUTLET

FUEL INLET —-B

 

 

 

—-w= BLANKET-SALY
QUTLET

 

 

 

BLANKET-SALT INLET —p-=—

Fig. l.'l.-2. MSR Layout with Fuel Pumps Equally
Spaced Around the Periphery aond the Blanket=Salt
Pumps in a Central Annulus of Blanket over the Core.

frame above the reactor. The fuel pumps are
connected by a gas annulus or torus with space
for expansion of the fuel under conditions other
than normal. The pumps ore completely covered
with fuel at all times, The fuel inventory for this
layout is 395 ft3 and could be reduced, since the
connecting pipes out of the core and the pump
tanks appear to be lorger than necessary. The
fuel inlet and outlet pipes pass through the
blanket pressure vessel wall and may create
undesirable stresses. The temperature difference
between the inlet and outlet pipes is expected to
be about 135°F. Further study will be required to
determine whether the resulting stresses can be
tolerated.

The layout shown in Fig. 1.1.3 has the five fuel
pumps ‘equally spaced around the top and the two

10

UNCLASSIFIED
ORNL-LR—DWG 27885

 

BLANKET-SALT
EXPANSION TANK —

 

 

 

 

 

 

 

 

T
FUEL INLET
FUEL OUTLET =S
/ // CORE \
[l o% |
|
2% \_j b
BLANKE
~-_ | "
-
L
\\
~~—3- BLANKET-SALT INLET

Fig. 1,1.3. MSR Layout with Five Fuel Pumps That
Have a Common Header and Offset BlankesSalt Pumps
That Draw from a Blonket Chamber Above the Fuel
Expansion Tank.

blanket-salt pumps offset in a blanket chamber
above the fuel expansion tank. The five fuel
pumps have a common header, The suction, volute,
and discharge of each pump are located in a trough
to minimize inventory and to direct the flow. The
fuel -returns through the fuel expansion tank,
gathers in a plenum, and is then directed to the
core. The fuel pumps are supported by their
barrels from the fuel expansion tank top. The
blanket salt flows upword around the core and
passes between the fuel pumps in rectangular

¥}

i

 
 

 

m

e

>y

"

W

UNCLASSIFIED
ORNL-LR-DWG 27886

 

  
 
 
 

L ANKET-SALT
PUMP
COOLANT
OUTLET

  

 

 

 

SALT
OUTLET

Fig. l.1.4. Single-Pass MSR L.ayout.

ducts to a plenum at the center of the blanket
deck from which the offset blanket-salt pumps

draw.. The blanket-salt: pumps .are suppor'red by;_-__
their barrels from #he blanket expansion tank top.'}_
'_fprobably be required. -
fuel. pumps. The fuei mventory for thls Iayout lsf :
’approxsmotely 440 f2, Sl e
. One of several prel:mmary onouts of smgie~pass
| 'r'fsystems is shown in Fig. 1.1.4. The single-pass -
- system” offers the posmblhty of closely ‘coupling
* the primary- heat exchangers to the reqctor.‘ Such -
~close couplmg would be of advantage in mmlmlzmg
" the size of the reactor ceil. ~Thermal sleeves are
_mcorporated to relieve - fhe sfresses | mvolved in

- connecting the outer blanket shell to the inlet und__'_"_i,
outlet piping. “The top of the reactor is somewhut;-

The reactor. supporfs ¢could be located ‘under the

simpler than fora two-poss system,- uithough some
of the complication was merely transferred to the.
bottom of the reactor. This layout required a fuel

\BLANKET-

 

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED
ORNL-LR-DWG 27887

 

 

 

BLANKET-SALT
PUMP

   

. BLANKET~SALT
. OUTLET

FUEL QUTLET
FUEL INLET

BLANKET- SALT INLET

Fige. L1,5 MSR Loyout with Fuel Return Ducts at
the Same Place as the Fuel Discharge Ducts,

mvenfory of 420 ft3, The reacfor support problem
was not consldered in these studies; however, a
toad ring ‘and pump bcrrel type of support would

"Another two-pass system is shown in Fag. 1 l 5

' The return fuel ducts are. brought back -to the top

_of the reactor at_the same place 'as ‘the pump dis-
charge ducts, (Results of flow studies of this

,;}’arrangement are _presented in. Chap 13) ~This

~feature tends to. Iower the over-all helght and

’,hence the pressure - dlfference across the core
_shell,  The plan view shows much fess blanket
'}_materlal in the - reglon of the dlscharge and return’
ducts. than there is in desngns having the return
- ducts - above ot below the discharge ducts. As in
all the ‘two- -pass” layouts with multiple pumps, the

top of the reactor is complex wn‘h regard to fabri-
cab:llty._

11

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

'NUCLEAR CALCULATIONS
L. G. Alexander J. T. Roberts

Nuclear calculations relative to parameter studies
of molten salt reactors for the production of power
were continued, The Univac calculations, for
which the Ocusol program is used, have provided
information about U235. and U?33-fueled reactors
of the reference design type, primarily, but studies
have been initiated of plutonium-fueled molten

‘salt reactors and of graphite-moderated molten salt
y233 breeder reactors.

7Unrivc‘lc-0cuso| Reactor Calculations
J. T. Roberts

The parameter study of two-region, homogeneous,
molten salt reactors described in the previous
report! has been extended and refined. The calcu-
lations based on the use of the 69 mole % LiF-=31
mole % BeF, salt plus ThF4 and UF , in the core
and of the 75 mole % LiF-25 mole % ThF, salt
in the blanket were completed, and a new study
based on the use of a 63 mole % LiF~37 mole %
BeF, salt plus ThF, and UF, in the core and of
a 71 mole % LiF-16 mole % BeF,~13mole % ThF,
salt in the blanket was initiated. The effects of
changing the salts are expected to be a small
reduction in critical mass and a reduction of
approximately 12% in the external regeneration
ratio. Blanket thickness will also be a variable
in the new parameter study.

In the calculations just completed a comparison
was made of U233- and U?35-fueled reactors. The
results of these calculations are presented in Fig.
1.1.6, which shows the fissionable material in-

ventory at startup as a function of core diameter.

The effect of various amounts of thorium, in the

“range 0 to 1 mole % in the fuel salt, is also shown.

For the some core diameter and thorium content of
the fuel salt, the U233 inventory for the cases
shown is less than half the U?3% inventory., This
is due to both the higher ¢, and the lower a for
y233 compared with those for U233 in intermediate
reactors, In Figs. 1.1.7 and 1.1.8, the 5-1, the
initial total regenercmon ratio, and the initial
blanket regeneration ratio for U235. and UZ33.
fueled reactors are plotted as functions of thorium
content of the fuel salt for core diameters in the

 

L. G. Alexander and J. T. Roberts, MSR Quar. Prog.
Rep. Oct. 31, 1957, ORNL-2431, p 5.

12

- UNCLASSIFIED
ORNL—-LR-DWG 27888

 

5000 —— 1T
SALT HOLDUP OUTSIDE CORE
ASSUMED TO BE 339 #3

 

 

 

 

 

 

 

- FUEL SALT: 69mole % LiF -3 mole %
BeF, + ThE, + UF,

BLANKET SALT: 75mole % LiF~25mole % ThE,

oo LN R

8

 

 

 

 

 

 

 

 

A 1N I
1 AN :
4 \\ R s U235, {mole % Th |—
IN FUEL SALT
N N

 

 

 

 

 

 

 

 

CRITICAL INVENTORY (kg of U3 o y23%)

200 ‘L\\\
\\ \: . U235, No Th

"k‘ ' U233, 4mole % Th

 

—]—_AU?35 025 mole % Th

N\
N
\ - L
\ T U235, 0.5mole % Th

 

100

 

U233 NO Th

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50

 

3 4 5 6 7 g 9 {0
CORE DIAMETER{ft)

Fig. 1.1.& Clean, Critical Inventory of U233 gnd '

U235 {or 600-Mw Reactors Fueled with Lithiume
Beryllium-Base Sclts.

range 5to 10 ft. In the U235 cases the regenera-
tion ratio is limited to a maximum of about 0.675
by the comparatively low value of 5 and a practical
minimum loss to capture in the salt and in the core
shell of about 0,1+, The U233 cases were not
carried for enough to establish a *‘ceiling’’ on the
regeneration ratio, but, judging from the n—1values,
it would seem to be possible to exceed 1.0 by
increasing the thorium content of the fuel salt to,
perhaps, 1.5 to 2% and, at the same time, in-
creasing the U233 inventory to, perhaps, 300 to
900 kg, depending on the core diameter. It should
be remembered, however, that the regeneration
ratios shown for the U233-fueled reactors will
decrease with time as burnup poisons accumulate
in the system, while in the U235.fueled reactors,
with initial regeneration ratios of about 0.6 or
better, the regeneration ratio may actually improve
with time and with reasonable chemical- process'ang
rates because of the buildup of the superior U%33
fuel,

Calculations were also made in order to compare
lithium-beryllium ond sodium-beryllium salts as

fuels for reactors of the reference design type.:

Figure 1.1.9 presents a comparison of the y23s
inventories at startup for reactors fueled with

U233, 0.25mole % Th ~—1-

L3

 
 

 

 

0

F.

P

P

) ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 UNCLASSIFIED
'ORNL-LR-DWG 27889
T T T T
FUEL SALT: 69 mole % LiF-31 mole % BeF,+ T, + UF,
41" BLANKET SALT: 75 mole % LiF 25 mole % ThF,
1.0
. \ CORE
0.9 DIAMETER
- \ \ 10 ft1
N
& 0.8 ™
: =
x \ N ("
© r— 5 ft
o , G ft
L 07
o —1_ ot TOTAL
z S| == 81t Y REGENERATION
S ' ft
2 os ?__‘e < et 2 1| 1 RATIO
& 6 /
% N/ A
g o0s %
|1
04 Ny —{5+#
. \ T 6 ft| |p
o3 . BLANKET |
' I REGENERATION
~——_ ~——_ RATIO
~—_ 8 ft
04
0 Yq s 3 4

THORIUM IN FUEL SALT (mole %)

Fige 1.1.7. _Initial Regeneration Characteristics of

Molten Salt Reactors Fueled with thhlum-Beryllium-

Base Salts Containing U235,

69 mole %" L|F—3I mole % BeF plus ThF and,'
UF  in- ‘the core and 75 mole % LiF-25 mole %
- ThF, in the blanket and for reactors fueled wnh
53 mole % NaF=47 mole % BeF, plus’ ThF and
,'_-{UF in the core and .58 mole % ‘NaF-35 mole % '
L BeF —7 mole % ThF in the bIanket. For-the same’
- core’ dsameter and thoruum confent of the fuel salt L
- the sodlum-beryllwm-base -salts requnred Llto.
2 times the U?3 mventory requured for the lithium-
beryllium-base salts. In Fig. 1.1.10 the 7-1,the -
- initial - total regenerahon rcmo, and - the mmql'
~ -blanket regenerunon ratio are ‘plotted for: reuctors'
' fueled with Na-Be salts, - In' comparison’ ‘with the
‘Li-Be-base " fuels (Fig. 1. 1.7), “the ‘Na- Be*base'fl_,g.

fuels show a disadvantage of 0.1 to 0.15 in
regeneration ratio. Since Na-Be salts cost ap-
proximately one-half as much as Li 7.Be salts,

_ concentmt:on of

mute, together with the

PERIOD ENDING JANUARY 31, 1958

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, UNCLASSIFIED
: . ORNL—LR—DWG 27830
1.3
| | _
o 0.25mole % Th
w2 b -NOTh - Pl _ ]1;— .
[ m—— imole % Th
£
10 _
7 \
= imole % Th
g \ \ TOTAL REGENERATION RATIO
£
a h
= o8 N _ !
g p \ \ _—025mole % Th
= N\ 25mole Y%
\ - h
567 \ - NS TOTAL REGENERATION __|
§ N, | ToTAL
T 086 , L REGENERATION RATIO
N (ALSO BLANKET
: ™. REGENERATION RATIO)
0.5 h —y i
\\ [
S /—0 25mole % Th
0.4 L BLANKET
’ <. REGENERATION |
- I~ RATIO
03 tmole ToTh—Sw_ | T~y
: BLANKET REGENERATION RATIO S
Y
] )
0.2 v
3 4 5 6 7 8 a 10

CORE DIAMETER (ft)

Fig. 1.1.8. Initia! Regeneration Charocteristics of
Molten Salt Reactors Fueled with Lithium-Beryilivne
Base Salts Containing u233,

some further studies of them will be made, even
though they do not compare very favorably on a
nucleor basm.

A comparlson of flux-averaged cfoss sections of
U235 U236, and U238 in reactors of the reference -
deSIgn type . indicated that - at steady state- the
Y23’ plus U238 will- be about
60 to 90% of the U235 concentration. This esti-
U233 critical concentration
“estimates, indicates that the use of 1 mole % UF
in corrosion tests is conservuhvely high if data

cppllcabie to the reactors of greatest interest are

" to_be obtained. The data may not, however, be

appllcable to. grophnte-moderated thermal reactors,

since the steudy-state ratios of even to odd uranium
“isotopes will be greater in thermal thon in mter-
medsate reactors. o

Cross sections for Ng, Be, Pu239 Pu‘uo C, ond
Bi have been added to the Ocusol sigma tape (for

13

 
 

 

//‘

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED
- ORNL-LR~-DWG 2789
5000 " T T T T 1 T T T

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Na-Be FUEL SALT: 53 mole % NaF-47 mole 7 BeFy
— PLUS ThF4 + UF4 -
No-Be BLANKET SALT: 58 mole % NaF-35mole 7 BeF,-
| 7mole % ThF,
Li~Be SALT: 69mole % LiF-31mole 7 BeFy PLUS
ThFg +UFy
. Ll-Be BLANKET SALT: 75 mole % LiF-25mole 7o ThF,
2000 |—  SALT HOLDUP OUTSIDE CORE ASSUMED TO BE
' 339 113
y o s o
~ * 0\
B : NN~ - y No-Be, {mole % Th
£\ N i
L 1000 -
o IO R
-:-‘: \ \ \\ \ \m\
S \ 1\ N g Li~Be, imole % Th —
> AV N
> \ . . N\
= : o
500 [—e- \ N
- \ \ T T~—g¢ Li-Be, 0.75mole % Th
VNN ]
| \ | ¢ Li-Be, 05 mole % Th
' | |
L y No-Be, NO Th
RN~
200 1 |
: 9 \“ . I ! ’
\ ~ & Li-Be,0.25 rpole % Th
\u.. b Li-Be, NO Th
100 . :
4 5 6 7 8 9 {0
CORE DIAMETER ({f1)
Fige. 1.1.9. Clean, Critical Inventory of U235 for

600-Mw Reactors Fueled with Lithlum-BerylHum- or
with’ Sodium-Beryllium-Buse Salts.

cross sections available previously see ref 2),
Also, thorium cross sections with o adjusted for
resonance saturation effects have been inserted
for 0.25, 0.5, 0.75, 7, and 13 mole % thorium in
" lithium-beryllium-base salts. Pseudo cross sec-
- tions “for combined- Li-Be-F in various core and
blanket salts have been inserted so that these
may be used as a single ‘‘element” in Ocusol
' culculatlons, whlch are limited to seven elements
in g reglon. '

dmcle.talcul-utions :
L. G. Alexander

" The Sorghum program was designed for computing
progressive ‘chonges in core concentrations and

 

= 2J. T. Roberts and L. G. Alexander, Cross Sections
for OCUSOL-A Program, ORNL CF+57-6-5{June 11, 1957).

14

UNCLASSIFIED
ORNL~LR-DWG 27892

 

 

 

 

 

 

-2 T T T T 1 |
FUEL SALT: 53 mole % NoF -47mole % BeF, + ThF, +UF4
1.| |—— BLANKET SALT: 58 mole % NaF - 35mole%Ber 7moie% ThF, — -
1.0 THORIUM IN FUEL SALT
| _A—qnoTh
0.9 — _
u/ : 2=
o8 tmole% Th
b tmole
— :

 

o
q

 

 

 

REGENERATION RATIO OR 7 -1
o
[

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

d 5 et tmole % Th]
0.4
TOTAL
b REGENERAT!ON
- ,
0.3 V\ RATIO
02 ' ‘\\ S —tro
T~ [ fhom BLANKET
O.f [~~——g {mole % Th 1| REGENERATION ___|
l , RATIO l
0

S 6 7 8 9 10
CORE DIAMETER (ft)

Figs, 1.1.10. Initicl Regeneration Characteristics of
Molten Salt Reactors Fueled with Sodium-Beryllium-
Base Salts Containing u23s,

regeneration ratios in molten salt reactors, with
Ocusol or Cornpone outputs being used for the
initial states, It is based on the assumption that
the scattering and leckage probabilities are in-
sensitive functions of the varying concentrations.
The program wos operated with constants that
characterized the spectrum, leckage, and scattering
in Ocusol Case 89-2, for which the data were pre-
sented previously.® The code operated satis-
factorily but disclosed the hitherto unrecognized
fact that the spectrum and leakage output from
Ocusol are not consistent with the scattering
properties. The source of the inconsistency is
not known, but may be due to a biased round-off of
the leakages at the highest energies, which, being
small, are known to only two significant figures.
Since the spectrum in Sorghum is computed in a
stepwise manner, beginning with the high-energy
groups, the round-off bias tends to accumulate
and to be amplified in the lower groups,

 

3L. G. Alexander and J. T. Roberts, MSR Quar. Prog.
Rep. Oct. 31, 1957, ORNL-2431, p 6, Table 1.2,

O

v

 
 

.

1

14

3

 

-

Y

 

“GM-GNU,” &

In order to remedy this defect, a subroutine for
computing the constants for Sorghum from the con-
centrations, ubsorphon cross sechons, initial
spectrum, and jeakage probabilities was written
and incorporated in the program. The resulting

constants, which differ only slightly from those

based on the scattering cross sections computed
by Ocusol, reproduce the initial spectrum and
critical concenfrahon exacfly.
Absorphon cross sections for Pa233, U234,
U236 Np237, and Pu?3? were estimated and were
inserted into the program, together with fission
cross sections for Pu?3?, The modafled program
has been operofed successfully on a test case.
Tests to determine limitations imposed on the
magnitude of the time ancrement by convergence
requirements and the drift in group leakage proba-

bilities are in progress.

Efforts to bring the Oracle program Cornpone
into use for molten salt reactors have been con-
tinved. Initial tests, for which the Ocusol group
structure and cross sections were used, showed
disagreement in the critical mass; also, the com-
puting time was excessive. The effects of modify-
ing the group structure, scaling factors, number of
space points in thin-shell regions, and method of
treating the principal scatterers are being studied,

A decision to normalize the fractional absorption
of neutrons in each elemental species with respect
to the reactor as a whole, rather than with respect
to the various regions, has necessitated revisions
of the edit subroutine. Provision has been made

to edrt the mtegrated flux -in each group and each L
region for use in Sorghum and other per?urbchon o

calculafrons.

IBM-704 Codes

and the *‘Curtiss-Wright 075," o two-dnmensnonui
multigroup, multiregion code. -

  

PERIOD ENDING JANUARY 31, 1958

FUEL FILL-AND-DRAIN SYSTEM DESIGN
G. D. Whitman

A preliminary design study has been made of a
fuel fill-and-drain system for the MSR. This
system must meet the following three major design
criteria;

1. A preheating system must be provided that is
capable of maintaining the drain vesse! and its
connecting plumbing at 1200°F,

2. A reliable heat-removal system must be pro-
vided that has sufficient capacity to handle the

fuel afterheat,

3. The drain vessel must be ‘‘ever-safe’’ so that
criticality cannot be achieved when the fuel is
drained. :

The fuel draining operation has not been con-

sidered as an emergency procedure which must be

accomplished in a relatively short period of time
in order to prevent a catastrophe, |t is considered
that, if all forced circulation ceased during. high-
power operation, the thermal capacity and heat
losses of the system would prevent prompt thermal
transients that would be capable of causing
failure of the primary containment vessels. There
are, however, other incentives for rapid removal of
the fluid from the fuel circuit. [f, for example,
there was a leak in the fuel system it would be

_important to drain the fuel in order to minimize the

cell contamination and cross contamination of

‘systems. Further, rapid removal of the fuel at the

time of a shutdown for maintenance would have an

‘economic advantage in reducing the power outage

'hme.
A consnderahon of these factors mdlcated that

e the maximum afterheat design load should be 10 Mw
" - for ' a 600-Mw reactor that had been operating.
'for one year- und had been shut down' for 10 min
e ~ before the fuel was drained.  No credit was taken
L D’ Baxter ST e g f'lssmn-gus removal during operation, - It was
The appllcablhfy of existing: lBM-704 codes 10 il
' molten salt reactor calculahons was - mveshgated
‘and three promising codes were selected for further -~
~study. These are: “the * CURE,”l a general:zed o
o ;,two-dlmensnoncl : mulhgroup program “suitable for -
ff_mvesngatmg the effect of ‘the north hecd of tfie-:;:_:"
- reactor on’ H1e reachvaty und breedmg ratio; the-;*-;l
“multigroup, . muihregton, -one-
dimensional program “similar to Eyewash-Ocusel; -

determined that 15 min would be reqmred to remove
f.'the fuel from the reactor. -~ » _

“For the drain vessel design calculuhons, it was

31-:\,ussumed that af 3200°F the fuel system volume
"~ would be 600 #3,~ “The design: cupacnty of the
‘drain vessel was’ therefore set at 750 #3 in order
- to ‘G“OW for jemperature excursuons and a residual
‘heel. An array of 12-in.-dia pipes was selected as
‘the primary containment vessel of the drain system

in order to obtain a large surface-area-to-volume
ratio for heat transfer efficiency and to provide a

15

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

large amount of nuclear poison material, Forty-
eight 20-ft lengths of 12-in.-dia pipe would be
arronged  in six vertical banks connected on

opposite ends with mitered joints. The six banks

of pipe would be connected at the bottom with a

common drain line that would connect with the

fuel system, The drain system would be preheated
and maintained ot the desired temperature with
electric heaters installed in small-diameter pipes
located axially inside the 12-in.-dia pipes. These
bayonet-type heaters could all be removed or in-
stalled from one face of the pipe array to facilitate
maintenance. The entire system would be installed
in an msuiated room or furnace to minimize heat
losses.

The removal of the fuel afterheat would be
accomplished by filling boiler tubes installed be-
tween the 12-in,-dia fuel-contdining pipes with
cold water from a header that would normally be
full.  The boiler tubes would normally be dry and
at the ambient temperature of about 1200°F,
Cooling would be accomplished by radiant heat
transfer from the fuel-containing pipes to the low-
pressure, low-temperature water in the boiler tubes.
For the peak afterheat load, about 150 gpm of
water would be required to supply the boiler tubes,

This fill-and-drain system design satisfies the
design criteria in that it is always in a standby
condition in which it is immediotely available for
drainage of the fuel, it can adequately handle the
fuel afterheat, and it provides double containment
of the fuel. The heat-removal scheme is essen-
tially self-regulating in that the amount of heat
removal is determined by the rate of heat transfer.
Both the water and the fuel systems are at low
pressure, and a double failure would be required

“for the two flvids to be mixed. The drain system

tank could be enclosed and sealed from the atmos-
phere because there are no large gas-cooling ducts
or other major external systems connected to it,
A stainless steel tray would be placed below
each bank of pipes to catch the fuel if a leak

- developed. These trays would be cooled by water

walls to prevent any possibility of meltdown and
destruction of the cell,

DESIGN OF A HYDROSTATIC BEARING FOR A
- MOLTEN SALT PUMP IMPELLER
| B. W. Kmyon

A hydrostatic bearing was investigated for use
as a pilot bearing for a molten salt pump because

16

the pressure of the pumped fluid could be used fo
maintain the centering of the impeller in the pump
casing.  Two designs are possible: one in which

the pockets are located around the .inside of the

bearing ond one-in which. the pockets are on the
impeller. The latter design offers the advantage
that the supply passages for fluid flow to the
pockets are in the impeller and are thus replace-
able. ' '

The chief difference between the conventional _

hydrostatic bearing with stationary pockets and
the bearing in which the pockets rotate on the

impeller is that in the latter case the fluid flow

through a given pocket is variable except when
the impeller is centered, which only occurs under

the improbable situation that there is no side-load

on the shaft. The variation in flow velocity, v, of
the fluid leaving the pockets of the two dessgns is
shown-in Fig. L.1.11,,

The pressure in the pocket achng between the
shaft and the journal is proportional to v2 if the
pockets are stationary, but the proportionality must
be adjusted by an acceleration factor if the pockets
are rotating. The curves of Fig. 1.1.11 indicate
that the load-supporting ability of the two designs

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
. ORNL=—LR—DWG 27893
1.2 :
o
X 10 7N
2 7N
© ROTATING  J/ v STATIONARY
S POCKET—7/ \\ POCKET
4 08 ! N\ '
a IMPELLER 4 \
3 CLEARANCE  {/; \
o CLOSING / )
L A \
o A “
& 06 ff IMPELLER
S i/ CLEARANCE
o y OPENING
> y
/
E 04 - X
1 .
w
o \\
0.2
0 60 120 180 240 - 300 360

ANGULAR POSITION OF POCKET {deg)

Fig. 1.L11.  Comparison of Velocities of Fluid
Leaving Pockets of Stationary and Rotating Hydrostatic
Bearing Systems,

r)

 
 

1)

i

$

.’\

 

 

o

is nearly the same, but there is a sllght angular

displacement between the curves. -

Preliminary calculations for a 16-in.<dia impeller

running at 1175 rpm ond having 12 pockets around

the periphery have been made. The rodial clearance
was taken as 0.015 in. The load capacity per inch
of pocket height is shown in Fig. 1.1.12 as a func-
tion of impeller dlsplacement. Ctrcumferenhal

flow between pockets was neglected in these pre-
liminary calculations. A calculation at one point

indicated - that the circumferential flow would re-
duce the load .capacity about 15% for a bearing
1.5 in. high.. |f the bearing herght were increased,
the relative circumferential flow would increase,
ond the load capacity of the bearing per inch of

henght would decrease, even though the total load

capacity of the bearing would increase, For this
bearing, a flow of 0.137 #13/sec was required with
a power consumption of 1.38 hp (net). These
values are. 1.25 and 0.82%, respectively, of the
flow and power consumption of the pump.

An investigation of the effect of '\./arying the

‘pressure and flow indicated that, for the same

power’ onsumphon, high flow at low pressure

would ‘give more load capacity in the bearmg than
would low . flow ot high pressure. The pressure

developed by o rotahng column of Huid is pro-

portional to Ro - R where R and R, are fhel

outer and inner radn of the flmd supply pcssoge

. PERIOD ENDING JANUARY 31, 1958

in the impeller, and therefore increasing R. and
enlarging the orifice would result in a larger load
capacity of the bearing without increasing the
power required for providing the flow for the
bearing.

UNCLASSIFIED
ORNL~LR-DWG 27894

 

 

 

 

 

 

 

 

 

400
£ 300
-~ @
o iy
- L
>
- Z
g&
x .
,%m 200 —
of : /
et | . |
o
Sz
@ o
W 100 /
.0 -
o 0005 0.0410 0.015

' IMPELLER DISPLACEMENT (ln)

Flg. LL12, Becring Ccpuchy vs Displucemenf of
Impeller Having 12 Packets Around Periphery and
Opercflng at 1175 pm.

17

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT -

1.2. COMPONENT DEVELOPMENT AND TESTING .

Ho w. chage ,

FUEL PUMP DESIGN AND DEVELOPMENT
W. F. Boudreau A. G, Grindell

Pump Design Studies
W. G. Cobb ‘M. E. Lackey

The optimum design of a fuel pump for a
circulating-fuel reactor is, of course, dependent on
the general configuration of the reactor and its
major components. In turn, the design of the
reactor can be substantially influenced by the
pump characteristics. This. mutual interdependence
dictates that pump design must be continuously
coordinated with the general reactor design.

One of the major considerations in the design
of a centrifugal pump is the problem of cavitation,
The onset of cavitotion is primarily governed by
the net total pressure available at the impeller
inlet to lift the salt into the impeller. Cavitation
must be avoided, since it can cause physical
damage to the impeller or to other parts of the
pump. - ‘Cavitation data obtained for fused salt
pumps at ORNL have been studied in relation to
pressures, flow rates, and pressure drops in the
heat exchangers and in other portions of the primary
fuel circuit of the reference-design reactor, These
preliminary studies indicate that five pumps with

the following design parameters will be suitable

for circulating fuel 130 (BeF,-LiF-UF,, 37-62-1

- mole %):
Flow through pump 4950 gpm
Heqd prcduced_ by pump 71 ft
Tempe’rature at pump inlet 1230°F ‘(max)
: Static pressure ot pump inlet 19.2 psia
Pump shaft speed ' 1160 rpm

o

W. Bo .MC,DOHOH .

A pump design is being prepared that is based on
these design values, but it is realized that further
studies of the reactor design may require changes.
It will not be difficult to translate the resulfs of
these studies to pumps of other ccpacmes.

The bdsic" design of an impeller and a double
volute for the pump wos developed that is serving
as the basis for a series of pump layouts. Various
types and arrangements of bearings (salt-lubricated,
oil-lubricated, and gas-lubricated) and seals are
being considered in conjunction with at least two
different motor arrangements (totally enclosed and

‘conventional). These layouts will permit a de-
termination of the most favorable arrangement of |

all elements from the standpoints of adaptability to

‘reactor construction, durability, adaptability fo

remote maintenance operations, reactor operutlonal
problems, and so on,

A number of commercial organizations have built
pumps for use with liquid metals under service
conditions which approach those of the molten
salt reactor. In order to utilize their experience,
study contracts are to be negotiated with several

- manufacturers. These companies will be asked

to study the fuel pump application in the fight of
their own experience and to formulate conceptual
designs that incorporate their recommendations.
Preliminary discussions have been held with
representatives of Allis<Chalmers Mfg. Co.,
Westinghouse Electric Corp., and Byron Jackson Co.

Development Tests of Solf-Lubticdted Bearings
P. G. Smith L. V. Wilson

The fused salt pumps now in use for development
tests have conventional oil-lubricated bearings.
These elements impose restrictions on the design
and construction of the pump, since the bearings

k)

 
 

1)

W

 

A%

)

- must be shielded from both the heat and the

radiation from the fuel salt in order to avoid damage
to the bearings and to the fubricant.’ The availa-
bility of a satisfactory bearing lubricated by the
fuel salt would eliminate these restrictions and
should permit improvements in pump construction
from the standpomts of reliability and sumpll-
fication.

A survey of extshng knowledge concerning

materials suitable for use in the construction of

salt-lubricated . bearings was made, and it was
found that at least two combinations of materials
showed considerable promise: (1) molybdenum
running against tungsten carbide containing
12.5 wt % cobalt and (2) tungsten carbide con-
taining 12.5 wt % cobalt running against itself,
The coefficients of thermal expansion of these
materials differ from those of INOR-8 and Inconel,
however, and therefore there may be design and

- development problems involved in the use of these

materials.

Two high-temperature beurlng test units are to
be set up. One unit will provide facilities for
testing hydrodynamic (plain journal) bearings, and
the other will provide for the testing of hydrostatic
(fluid piston) bearings. Existing high-temperature
equipment »wili be modified-for use in this test
program.

The design of the hydrodynamic-bearing test

~ unit has been completed, and procurement and

fabrication of parts have been initicted.  This

test unit will utilize a modified sump pump
designed for circulating NaK at high temperatures.
The impeller has been removed, and the shaft has -
- been modified to accommodate the ‘rotating portion ..
of the test bearing ot its lower end. The stationary =~ -
© - element of the bearmg is held by a movable arm
'.;_broughf ‘out: of the pump tank through a bellows
- seal. Vonuble loads can be apphed to the: beanng‘ '
. "by ‘andir. cylmder attached to the outer end.of
- the movabie -arms A callbrated drive motor is used
" in order to obtain data on load vs test-unit torque.
-~ ~The .test’ mformohon ‘that “ will be - obtained is
. _expected to permit a good analysns of the’ behuvior -
of the fused salts of interest-as bearing lubricants.

-~ The theory of the hydrodynamlc bearing has been
Do -well ‘worked out in recent years, and its relative
" ‘mechanical snmplncity tends to facilitate analysis.
“The initial ‘studies -of - the pump design problem:
have indicated, however, that. the hydrostatic’

(fluid-piston) type of bearing may have a number of
advantages for use in fused salt pumps. Among

PERIOD ENDING JANUARY 31, 1958

the most important is thar such bearings can
withstand a considerable amount of mechanical
damage without appreciable impairment of their
load-carrying capacity.  Therefore the second
bearing -test unit will incorporate a hydrostatic
bearing. Layout studies indicate that a satisfactory
test unit can be constructed by removing the lower
(oil-lubricated) bearing and seal of another pump
of the type described above and incorporating a
hydrostatic bearing on-the shaft immediately above
the impeller. The impeller will be used as the
source of pressure and load for the bearing.

The novel form of hydrostatic bearing described in
Chap. 1.1, which eliminates complexities associated
with many conventional hydrostatic bearings, will
also be tested. A model that will use water as the
lubricant is to be constructed.

" The combined geometries of the reactor and the

pumps will, of course, determine eventually the
location of the pump bearings. Since salt-lubricated
bearings could be placed at several locations in
the pump, studies of a variety of configurations of
salt-lubricated bearings are being made in terms
of the basic variables, such as diameter, clearance
ratio, and length-to-diameter ratio.

| ‘Development Tests of Mechanical Shaft Seals

i.Development tests of sump pumps with oil-

“lubricated shaft seals that were designed for
circulating high-temperature NaK have revealed

that NaK vapor may diffuse into the seal oil-
leakage area and react with the oil to form sludges

_that__interfere . with the functioning of the seal.

Smce onl lubrlcafed shaft seals would be of use in -

i':_r‘molfen ~salt reactor appllcahons, means for-,
“preventing diffusion of process-fluud vapor into
' _:the seol region are belng sought, :

The flow pafh of the helivm in the seal area

"'-_'.(wh|ch normcfly serves to purge the oil leaking
“from  the seal) was changed so that port of the
~helium would flow upward into the seal area ond
~ the remainder would flow downward along the shaft.-

to reduce dlffusson of the NaK vapor into the seal -

'“"feglon. *.A pump mcorpomtmg this modified seal

has now operated for upproxnmctely 2600 hr at o

. NaK temperature -of 1200°F and o pump speed

of 2400 rpm. Seal leakage has been negligible,
and no evidences of seal damage have been noted.

- The test is continuing.

19

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS kEPORT.

Develepment' Tests of OilsLubricated Be_cr.ings(
P. G. Smith . W. E. Thomo_s

The lower bearings for the fuel b'umps'_ may :

possibly be salt-lubricated, as described above,

but oil-lubricated bearings may be required at the
‘upper end of the shaft. The oil in these bearings

could, of course, be damaged by radiation from the
salt fuel. It has been found that the lubricant

Dowtherm A, o eutectic mixture of diphenyl ond .
diphenyl oxide, offers the advantage that it
decomposes into gaseous and liquid products when

irradiated rather than into carbonaceous deposits

hydrocarbon<base lubricants. The principal dis-
advantage of Dowtherm as a lubricant is that it

lacks the property of oiliness, that is, the ublllty

to provide very thin films for lubrication. -

~ An experimental evaluation of the bevh_cvio'r of
Dowtherm A is being condutted in a special high-
temperature sump pump that includes centrifuges

for degassing. The pump is circulating fuel 30
(NaF-ZrF4-UF4,"50-46'4 mole %) at a temperature
of 1200°F; the shaft speed is about 3000 rpm; and
the temperature of the lubricant returning from the
bearing is maintained at 180°F.

The initial test was terminated by failure of the

drive motor after 172 hr of operation. Future tests
will-be on a long-term basis, except that the pump
will be dismantled in order to inspect it for signs

of wear at intervals of 1000 hr. Aside from the
swelling of some O-rings made of buna N elastomer,

no adverse effects attributable to the use of
Dowtherm Aasa iubrlcant have been noted

| Irtqaiufioh and Endurance Tests of
 Bearings and Seals

S. M._ DeCamp - W. E. Thomas

Bearings exposed to radiation fields could be

lubricated with hydrocarbons if it were possible to

limit the dose received by any part of the lubricant
to .a maximum value of about 10 r. In theory, .
this. could be accomplished by changing the.
~ lubricant frequently enough to avoid an excessive
tofcl dose. In practice, it would be difficult to be

certain that all the lubricant was: ‘temoved from the
area of greatest exposure in a sufficiently uniform

manner. Irradlctlon tests have . demonstrated that -

any porflon of a hydrocarbon lubricant that receives

an excessive dose of radmtlon will form solid -
- resrdues and (for some types of lubnccnt) OCICIIC

20

products. Such reaction products could be expected

-to interfere with the removal of the balance of the
lubricant from the irradiated area.

A test unit for investigating this problem in a
practical situation has been set up in the canal ot
the MTR. Spent fuel elements are placed around

~ the seal and bearing area of a test unit composed

of a bearing housing, .a (lower) sleeve bearing, an

- (upper) double-row ball bearing, upper and lower
- shaft seals, and a shaft.. The lower seal and
“bearing area will be exposed to o total dose of
1.3 x 1010 rep of gamma radiation. Test operatlon

- of this unit will be started soon.
‘of the type that result from the irradiation of

“A life test of a bellows-mounted seal (oll metal,

“‘except for the carbon nose piece) is bemg
" conducted in a sump pump that is circulating
" NaK at 1250°F; the pump shaft speed is 2400 rpm.
| .Alfhough the |eukage of the seal was high when

it was first installed, it has been negligible during

‘the last 400 hr of an accumulated total of 2500 hr

of operation.

DEVELOPMENT OF TECHNIQUES FOR REMOTE
MAINTENANCE OF THE REACTOR SYSTEM

E. Storto

The cbuln‘y to perform maintenance work on the
primary and ‘secondary systems of the reactor
within the containment cell is of primary importance
to the successful operation of a central-station

_nuclear power plant. Consideration of the problems
‘involved indicates that the most feasible solution

to this maintenance problem is to make all com-
ponents - removable and replaceable by remote

“manipulation and to arrange and interconnect the

components so that the number of replaceable units

"is minimized. Actual repair work on or disposal

of components that had failed would then be

carried out in separate hot-cell facilities set up

for ‘this purpose. The. problems associated with
this maintenance concept have been studied, and’
solutions now being worked out are described in

~ the following sections.

‘Mechanical Joint Development
In order for all components of the reactor to be

removable and replaceable by remote manipulation,
it is necessary to provide a means for remotely

‘separating and ‘joining the pipes which “connect
“the major componenfs of the system. Welded joints
- provide a system of the highest" ‘integrity, of
‘course, “and ‘methods for remote - ‘welding and

 
 

 

 

 

£

inspection are being developed -at other instal-
lotions. The difficulties are formidable, however,

and satisfactory equipment and techriques cannot
_be expected in the near future. Attention is

therefore ' being given to the development of a
reliable mechanical joint.

The freeze-flange type of joint shown in

Fig. 1.2.1 is being tested in the Inconel loop

illustrated schematically in Fig. 1.2.2. In this

type of joint, a frozen seal of molten salt is used
to prevent leakage of the salt from the loop through
the space between the flanges. An additional
seal of soft iron or copper rings is provided to
minimize gas leakage during shutdowns of the loop.

Heat transfer from the molten salt in the tubing

to the outer edge of the flange is reduced by
providing a narrow section in the flange, In
addition, there is a channel for air cooling in each
flange. The test flanges are made of Inconel and
are held together with eight quick-opening toggle
clamps, as shown in Fig. 1.2.3.

UNCLASSIFIED
~ ORNL-LR-DWG 27895

-—“~GAP("'/l6 in}

,-SOFT- ERON OR COPPER SEAL RING

—AIR CHANNEL FOR COOLING

FROZEN-SALT SEAL

 

 

 

L1 /—WELD OF FLANGE TO LOCP TUBING

: --F— SALTFLOW

RING INSERT TO PROVIDE LABYRINTH
“- - FOR SALT LEAKAGE :

   

 

 

 

 

 

 

 

- VNARROW secnow T0 REDUCE HEAT
| TRANSFER FROM THE MOLTEN SALT -
. mme LOOP TUBING

 

 

 

 

 

. INCHES - -

— Am_mu_t-:r‘fdcoon.mc c_'H'ANNEL}_.

: BETWEEN FLANGE FACES INDiCATE REGION OF FROZEN =SALYL

 SEAL ; E222 INDICATES REGION OF TRANSITION FROM LIQUID -
o 'TOSOLIDSALT. . .- S

Fig. 1.2.). Diagram of Freeze-Flange Joint.

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED
ORNL-LR-DWG 27896

‘ E _—FLANGE B

FREEZE-FLANGE JOINTS
(AXtS HORIZONTAL}

FLANGE A
HEATER CONNECTION LUG
| LOOP TUBING~INCONEL,
1/ ~in. OD, 0.045~in. WALL
SALT FLOW
CAST-METAL-SEALED : {:] _
FLANGE JOINT ADDITION

TO LOOP

SALT PUMP

 

 

 

  
  

 

 

 

 

‘ e ‘
{. CAST-METAL-SEALED
: | : FLANGE JOINTS{AXIS '
lL © VERTICAL) :
o . e —————
SALT FLOW

Fig. 1.2.2. Diagram of Loop for Testing Freeze-Flange
Joints and Cost-Metal-Secled Flange Joints.

 During the loop tests, temperatures were

‘measured ot the points lndlcated in Fig. 1.2.4.

Fuel 30, which has a melting point of about 970°F,
was citculated in the loop at a flow rate of 2 gpm.
In the first test the salt temperature was cycied
50 times between 1150 and 1300°F, and then the

loop was operated isothermally for several days

at 1300°F. The average cycle time was 44 min;
a cycle is defined as a temperature variation from

1300 to 1150°F and back to ]300°F. In the second
.- ‘test the temperature vcriohon of the salt was from

'.3‘1300 to ‘1500°F and back fo 1300°F, and the

= _'FROZEN -sALT SEAL

number of cycles was reduced to 30. The average

- cycle time was 29 min. - Representative tempera-
" tures measured durmg both tests are listed in

Table 12,1,

Both joints operoted successfully, there was. no

_indication of salt leckage. Neither the flanges nor
‘the seal rings were damaged by the high tempera-
 tures or the strains resulting from the thermal
- cycling. . The frozen: seals found upon separating.
~ -the flanges may be seen in Fig. 1.2.5, and a flange
. from which the salt has been removed is shown in

Fig. 1.2.6. Prior to taking these photographs, the
salt had been drained from the loop and the loop
had cooled to room temperature. -

21

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

A new freeze-flange joint is being installed in

the loop. The new flanges are identical to those -
described above except that copper sealing rings

are used in place of iron rings. A similar freeze-
flange |omf is being designed for use in a 4-in.

plpe.

UNCLASSIFIED|
PHOTO 204 |

s

 

Fig. 1.1.3. Freeze-Flange Joints In fesf téop.

UNCLASSIFIED

ORNL-LR —DWG 27897

  
 

16

15
14

— FLANGE B

—=— SALT FLOW

 
 
 
 

 
 

- 23
9 22
' 21

 
   
   
  

35
26
27, FLANGE A
28

 
 

Fig. 1.2.4. Locations of Thermocouples on Freeze-

Flange Joints.

22

 

Gas leakage tests made at room temperature with
a helium leak detector show leakage rates for these
1omts that are below the specfi:ed maximum rate of
10~7 cm3/sec. One joint had a leakage rate of

3 x 107 cm3/sec, ond the rate for another was

2.4 x 10-8 cm3/sec.

A flange joint which has ¢ cast-metcl seal
between the two halves has also been designed,
and tests are to be made with two sealing ma-

‘terials: silver and an alloy of 72% silver and 28%
_copper, (see Chap. 2.1 for details of selection of

sealing materials). A seal will be formed by

fusion of the sealing material before the loop is

chorged with salt. The seal material will be in
the solid state during operation. Joints of this

“type are being |ncorporoted into the test loop, as
~shown in Fig. 1.2.2. A diagram of the cast-metal-
- seal flange joint is shown in Fig. 1.2.7.

A flange joint with o soft-annealed copper ring

for sealing the two flanges has also been designed.

A raised tooth of Y-shaped cross section, machined
on the face of each flange, indents o flat copper

;. UNCLASS|FIED
; PHOTO 36205

Fig. 1.2.5, .Flun.g'es S_hov;n in Fig; 1.2.3 After Tests.

Note frozen-salt seals.

 

 
 

£d

 

) : .

 

 

iy

B :_Tdbla‘ To2.1. Freeze-Flange Temperatures Measured During Th'er'mnl‘ Cycles

 

' ‘.'I"e‘s‘f N‘o.‘l'- Slultv-Temperufure Cycled 50 Times

_Between 1150 and 1300°F

Test Nos 2 — Salt Temperature Cyclad 30 Times

Between 1300 and 1500°F

 

 

 

'Thermfi;:dfih'e" _ — . o = .
Number* . " ‘Minimum . Cyelo Mnximum‘. Cycle Minimgm Cycle - Maximum | Cycle
J .-Te.mp:rut‘qre ‘Number Temp: rature Number Tempoe roture Number Temp: rature Number
(CF) Lo (CF) ("F) ‘ ("F) o
21 | 862 6. 1018 33,50 960 3 1170 21,24
2 625 16 760 33,50 695 L3 950 9
23 450 6,14,20 577 16 500 3 652 9
24 245 6 280 50 275 3 305 30
25 693 0 820 4 810 6,12 942 1
26 490 16 580 4 582 6,21 665 )
27 348 16 405 50 a10 6 452 1,9,27,30
28 205 16 240 4 250 13,6 260 9,24
13 798 6 957 33 915 3,6 N2 2
14 560 16 685 a1 660 13 905 27
15 375 16 447 4 465 6 595 15,21,27
16 200 . 1006 228 37 255 24 305 27
7 748 16 865 2 865 - 12,1821 1010 1,3
1 510 16 590 2 615 12,8 700 16
19 315 16 362 50 410 6 450 3,6,9,30
20 185 16 215 50 260 . 21 278 "

30

 

*See Fig. 1.2.3 for ‘Iocution of thermocouple.

8S61 ‘1€ AYYNNYI 9NIGN3 Q0o1d3d

 

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFHED
2021

Fig. 1.2.6. Tested Flange After Removal of Frozen Salt.

UNCLASSIFIED
CRNL-LR-DWG 27898

 

SEAL MATERIAL INSERT
SHOWN BEFORE BEING
FUSED TO FORM SEAL

 

 

 

 

 

 

 

-WELD OF - - / INCHES
FLANGE TO LOOP TUBI_NG .

Fig. 1.2.7. Diqgrufi of Cust-Meful-Sealed Flonge Joint. -

™

 

- ring"’to ‘rprov'ide a tight seal when the joint is
formed. A diagram of this indented-seal flange is

shown in Fig. 1.2.8, an_d'a_ctucl parts are shown in

7 _ Flg- 102 90 ‘

Gas leakage tests mcde ‘at ‘room temperarure
with o helium leak detector showed the indented-

seal. flange joint to have o leckage rate of
1.3 x 10=% cm3/sec when. the two flanges were

~ clamped together - with a new copper seal ring,
A leakage rate of § x 108 ¢m3/sec was obtained -
" after reqssembly of the flanges without replucemenf

of  the prewously indented copper ring. . Both

‘leakage rates ‘are below the leakage rate allbw'ed_:
by the specifications. ngh-temperuture tests of

this type of |omt are plcmned

UNCLASSIFIED
. ORNL-LR-DWG 27899

 

 

 

 

 

»~ " LOOP TUBING -

| o 4 =7 o — -

 

 

 

 

ANNEALED COPPER RING - NCHES

Fig. 1.2.8. Diagram of Indented-Seal Flofi.ge Joint.

Remote Manipulation Techniques

C. K. McGlothlon

Consideration of the various types of remote<
handling apparatus now cvallab!e commercnc”y

indicates that they may be applicable to mamfe-_

nance of the molten salt reactor system. Ex-
“perience in the use of such apparutus has indicated,
however, that some modifications of the commercial - -
equipment would be desirable and that considerable
work will be required to develop special tools -
and techniques for using the eqmpmenf effechvely.
" Equipment has been set up for ‘studying the

problems associated with . remote ' maintenance

‘work. The component being used for the initial

experiments is a pump that has -been‘ruse_d for
circulating NaK at high temperatures. The manipu-.

lation apparatus is a new General Mills model Ef3

_WELD OF FLANGETO -

\\ RAISED TOOTH - _2 0 Y%
e —

 
 

 

&

D

#i

H

 

 

PERIOD ENDING JANUARY 31, 7958

 

 

 

 

 

 

 

 

 

Fig. 1.2.9. Components of an Indented-Seal Flonge Joint.

mechanical arm mounted on the standard General
Mills carriage and traveling bridge. The opercmons
to be performed are hsted below: ,

1. Remove the pump cover flange bolts.

2. Disconnect the tachometer cables. -

' 3. Disconnect oil and cover-gas service lines.
4. Insert the lifting eye in the impeller shaft, -

5. Loosen the pump cover flange.’
6. Remove the pumprotary element. =
- 7. Align and replace the pump rotary element,
8. Replace the : ‘pump cover flange bolts.
9. Tighten the pump cover flange. -
10. Reconnect the service lines.
‘11, Reconnect the tachometer cables.

Moving plctures will be taken to uss:st in motion -
studies ond - in fhe ‘development  of techmques J

applicable to the handling of other components.
The cell in which the experiments are to be
performed is shown in Fig. 1.2.10.

\ . 3. hinged

Heater-lnsulation Units for Remote Application

A. L. Southem

AH plpes and components of the reactor system

--_Wl” have to be preheated and insulated. Therefore

methods are being studied for rapid, remote removal

- and replacement of heater elements, and prelimincry
“studies have indicated that heater-insulation units
“-are _required that wrll tncorporafe the followmg
','_’"feutures. ’ :

e short, light sectlons contoured to fst ‘specific

,components or pipe sectlons,

2. ceramic clamshell heating elements built |nto

~ the msulahon blanket,
construction to - permit unfastenlng,

‘removal, lifting, und replacement as a unit,

4. plug-in electrical connections.

Design work on such heater-insulation units has
been initiated. Vendors of resistance-heating

25

 
 

 

 

 

 

'MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Fig. 1.2.10. Cell for Remote Maintenance Experiments.

elements have indicated their willingness ta.assist

‘in developing heaters to meet the requirements,

and plans -are being made to build and test
experlmenml unlts. '

Wintéhc-flée'Demonsfl'ofion Fa:ilfty |

E Storto

, The feus:blhty of system features, ‘equipment,
- and technlques developed experimentally for
; upphcoflon ‘to g remotely maintainable reactor . .

sysfem is to be determined in a large-scale demon-

~stration facility. The configuration of this facility.
will be essentially the same as that of the reactor.
cell, and mockups of components will be used to
' ,'determlne the best arrangements for. occessnbihty,
‘minimization of fuel holdup and pumping power,
- ease of observation, and efficiency of space

o

utilization. Equipment will be provided for circy-
lating o barren fused salt at reactor design”tenipér'-

atures and pressures in simulated reactor primary - ...
and secondary circuits in order to test mechanlcal";f,f;fl_'-_ Chi
joints, the reploceab:hty of componenfs, ’rhe';', S
adequacy of heater-insulation unlfs, the unmzahon l _ B
of wiring harness and service piping,’ and the_fr"_'._'-'"*""
application of - remote-wewmg an& -hundlmg

apparatus and techniques.

“Thus far, preliminary facrh.ty layout studles have’v,,f’
been made, and work has been’ sfarted on‘a sma”".,’_‘j- Core A

scale wooden and plastic model o -assist

“visualization and resolution of space confhcts.'_'{-;
Preliminary specifications have been prepared for =~ = -
~ the  procurement of a remote mampulafor of the__'f?-__, fi
General Mills model E-3 type .that” mcorporates S

modifications - suggested by . ORNL' ‘experience

with  this eqmpment. Work orders have been

 

 

 

 
 

oo

 

b}

 

9

written for the clearing of an areg for the facility

~in Bldg. 9201-3 at the Y-12 Plant “and for con-

struction of a craneway for the mqn;pylator bridge.
Arrangements have been made to‘study the appli-
cation of a commercial, closed, television circuit
to reactor systei-n maintenance problems.

HEAT EXCHANGER DEVELOPMENT

J. C. Arn.:os R. E._MacPherson
R,' l.. Senn

Heat transfer correlations for use in predicting
heat exchanger performance are required for the
development of a realistic heat exchanger design.
While standard heat transfer correlations have been
found to be. satisfactory for predicting the heat
transfer performance of some molten salts, the
performance of some other salts has been found to
deviate from'these correlations for various reasons,
For excmple, the . heat -transfer characteristics of

some salts appear fo be critically affected by the

type of material selected for heat exchanger

construction. Also, differences in heat transfer

performance of some salt mixtures have been
attributed to a difference in wetting properties of
the salts, - It ‘is, therefore, essential that heat

~ transfer correlations for the specific salt mixtures

under consideration for use in the - molten salt
reactor be determined experimentally so that
proposed heat exchanger de5|gns can be adequately
evaluated,

Heat transfer coefficnents will be determined

in laboratory-scale single-tube test apparatus

(see Chap. 1.3 of this report), and it has al-

 _ready- been possible - to- obtain pertinent fused
salt heat transfer data from a relatively Iarge-scale
~ heat exchanger test facility that was available,
A schematic flow dlagrcm of -the. test facility is . .
. shown in Fig. 1.2.11,  All ‘components of this
" system are Inconel, and heat is supplled to the
““fused salt by dnrect resistance heating of the
" Inconel tubing. = The heat -is transferred from -
. the- salt flowing in the ‘tubes to NaK flowmg n
_ “the shell side of a 25-tube heat exchanger of the -
. type. shown in Fig. 1.2.12, and heat is removedf _
~ from the NaK by a 500-kw NaK-to-cur heat dump. -
" Salt and NaK flow rates are measured by venturis
T «eqmpped wnh Moore Nullmnnc pressure-measuring -
' 'dewces. The NaK flow rate is also measured with

an electromagnetic flowmeter. All temperatures

"are measured with Chromel-Alume! thermocouples.

The NaK loop is equipped with a circulating cold

PERIOD ENDING JANUARY 31, 1958

trap for control of the oxide content. This test
facility, which is designated Small Heat Exchanger
Test Stand.C (SHE-C), is shown in Figs. 1.2,13
and 1.2.14. “Heat transfer data are presently being

obtained for fuel 130. Other fuel salts will be

tested ‘as required.

The system was calibrated by making a test run
with “fuel 30 (NaF-ZrF ,-UF ,, 50-46-4 mole %),
for which heat transfer data were available. Data
were obtained over a fused salt Reynolds number
range of 425 to 6150 ot temperatures ranging from
1100 to 1300°F. The heat balances obtained were
in good agreement throughout; there was, normally,
agreement to within 2%.

Preliminary ~analyses of data obtained for
fuel 130 are being made. Since the physical
property data currently being used for this salt
were extrapolated from data on other salts, com-
plete correlation of these data with results for
other fluids cannot be made until physical property
data for fuel 130 are determined experimentally.
Approximate  heat transfer coefficients can,
however, be obtained with the current information.
A block diagram showing operating conditions for
a typical heat transfer run is presented in
Fig. 1.2.15. '

The difficulty of accurately predicting liquid
metal heat transfer performance in heat exchanger

" shells may moke desirable the further refinement
. of the fused salt data by substituting a fused salt

of known heat transfer properties for the NaK now
being circulated in the secondary loop. Such a
substitution, if made, would require further
modifications of the test facilities.

: " DESIGN, CONSTRUCT!ON AND OPERATION

OF MATERIALS TESTING LOOPS
Forced-Cnrculafion Loops
- J L. Crowley -

The corrosnon of various materials, particularly

‘Inconel and INOR-8, by fused salt mixtures in the

temperature range 1100 to 1300°F is belng investi-

gated with- forced-c:rculahon loops in whlch high

th.ermql_ gr_ad:ents._con be attained in order to
simulate - reactor -operating conditions. ~ Low

_ corrosion rates are -anticipated for “Inconel and
~ INOR-8 in contact with the fused salts.of interest -

in the MSR program, and thus fong-time effects
are to be investigated. Satisfactory operation of a
loop will imply operation for one year or longer

27

 
 

 

 

 

 

 

 

 

"UNCLASSIFIED
ORNL-LR-DWG 27900

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~ NoK PUMP ~ SALT PUMP
' ELE%F\;?T:EGTNEERT'C THERMOCOUPLES
3 VENTURI
]
VENTURI )l Y
-' - PRESSURE ~MEASURING .
| DEVICES Y PRESSURE-MEASURING [ )
l . C DEVICES —
500-kw | «—RESISTANCE
NaK-TO-AIR | HEATER
RADIATOR | ECONOMIZER
. HEAT
‘ <c:0|_D TRAP VALVE EXCHANGER ]
[ J== ELECTROMAGNETIC
> FLOW METER
3 :
COLD TRAP | - ( J.._._ i
‘ . ' \/
NaK FLOW SALT FLOW
Ifi THERMOCOUPLES | I{I
— QPRESSURE—MEASUQING DEVlces'/_) — -
~ NoK - . o SALT
- SUMP TANK SUMP TANK

 

 

 

 

 

 

Fig. 1.2.11. Schematic Flow Diagram m‘ Small Heat Exchanger Test Stand C. o

L¥0dI Y $53490dd WYII0Id ¥OLIVIY LIVS-NILTOW

 
6¢C

   
 
 
   
  
    
   

107in. {REF}

 

MATERIAL: INCONEL . :
S 26 TUBES (0.1875-in. OD, 0.025~In. WALLS}
FUSED SALT OUTLET

 

FUSED SALT INLET

   

SECTION A-A

FUSED SALT HEADER

HEAT EXCHANGER TUBES

| NoK HEADER—

   

'HEAT EXCHANGER SHELL

SEfS COMB SPACERS .
SPACER WIRE 0.03% x 0.055 in.

UNCLASSIFIED
ORNL-LR-DWG 15289R

 

ISOMETRIC OF HEAT EXCHANGER

    

_ NaK INLET

| .,‘Fi;q..’ 'l‘.2‘.'l2. Diogffim. of Small Heat Exchanger of the Type Being Used To Obtain Fused Salt Heat Transfer Data.

8541 ‘L€ AYVNANYF 9NIGN3 Q01¥3d

 

 
 

 

 

 

 

i
j
1
i
i
|
I
I
i
1
i
|
[
i
{
|
i

30

 

 

RESISTANCE _
" HEATER

Fig. 1.2.14. Piping Arrdnigemerfl. of Sma

 

"Nak-TO-AIR RADIATOR 3

H Hedf'Exchu.nger Tést Stand :C.'

 

 

PHOTO 24914

 

 

 

 

UNCLASSIFIED

 

 

 

 

 

 

 

 

  

 

 

 

-

 
 

L]

&

S

 

UNCLASSIFIED
& éfiNL—LR DWG 27904

 

NoK-TO-AIR
| RADIATOR

 

 

 

 

NaK PUMP

NaK FLOW, 18,600 Ib/hr
. Rttt ——————————————r
H52°F 944 °F
36.0 psig FUSED-SALT-TO-NoK 106.0psig
1246°F |- HEAT EXCHANGER 06 °F

94.2 psu; 20.25 psig

o

FUSED SALT FLOW 15,900 Ib/hr

 

 

 

      
  

 

 

 

\ FUSED
SALT
PUMP

 

RESISTANCE
HEATER -

 

 

 

 

 

280 kw

Fig.fl 1. 2.15  Block Diugrutn Showing Typical Oper;
ating Conditions for @ Hent Transfer Run in Test Stond
SHE-C.

without significant equipment difficulties or
changes in operdting conditions. The scope of the
planned tests requires at least 15 test facilities,
nine of which were available at the start of the
quarter and two of which became available during
the period. The remaining loops are being
constructed.

The operating experience thus far has shown

the need for improvements in test stand rellobthty, :

and therefore more precise methods for inspecting,
maintaining, and assuring continuity of operation

fnconel and three INOR alloy loops are c:rculatmg

- salt mlxtures, and one lnconel ‘bifluid- loop ‘s in. -
operation - that has separqte sodium “and  salt
_ circuits connected through a U-tube heat exchanger.' :
AH  the loops - ‘ceased to operate and. the salts.
“froze during an_ unanhctpcted unprotected power - _
plates hold’ the tods in place and distribute the
flow. - The - ,5-m. rods are. Separated from each
other. by wire ‘spacers wound ‘around them at 3-|n._; o
lmtervais. The rods ‘were corefully weighed after - -

_the 'spacers were . mstaHed ‘and their- posmons

outage, and two Inconel Ioops, mcludmg ‘the =
~ bifluid-loop, failed as a result, - For the INOR-8_
_ Ioops fabrlcated ‘to dute, Husteiloy B pumps -have
~ been used  because. INOR-§- “pumps were not -
. -,ovallable., “Three of these. iNOR 8- loops - hove:..ij_: |
* failed in the sections made of Hastelloy B, which' ~-w
does not weld readily and becomes brittle, INOR-8
pumps will be used in future loops. (Discussions -

of solutions to fabrication problems are presented

‘in Chap. 2.1 of thls report.)

PERIOD ENDING JANUARY 31, 1958

The ovatlablltty of good quality INOR-8 is
improving, 'and two special INOR-8 loops have
been constructed. One of these includes graphite
specnmens and the other includes specml INOR-8
specimens for weight loss studies.

In the weight loss studies an effort will be made
to determine accurately the corrosion rate of
INOR-8 at a wall temperature of 1300°F, when
exposed to a fused salt, by carefully measuring
and weighing the inserts. Provisions will be

- made to assure that the specimens are not exposed

to the atmosphere before and after exposure. Three
samples will be exposed, each for a different time,
in an effort to establish a corrosion rate which
may be extrapolated to operating periods of severol
years.

The loop for the weight loss studies was
designed, as shown in Fig. 1.2.16, to incorporate
the three samples at the hot end of the heater
section with a separate power source to control
the wall temperatures. The samples were machined
to fit inside a sleeve. The samples.and the sleeve

~were fused together at the ends when welded into

the loop piping so that current would pass through
the samples.

The loop which contains graphite specimens will
be examined after operation to determine the

_extent to which graphite causes carburization of

INOR-8 aond the effects of the fused salt on
graphite,.  Specimens of impervious graphite

‘especially prepared by the National Carbon
- Company are being used.” The graphite-specimen

assembly is shown in Fig. 1.2.17. The graphite
container was designed to give a graphite-to-

of test stand components are being provnded. Four - INOR-8- surface area ratio of 0.67. It has been
 installed at the ‘outlet of the heater section of a

stqndard forced-cnrculatlon “foop - fubr;ccted of

/-ln.-dla, _ 0.035-|n.-wa|| INOR-8 tubing. "The

‘ gruphlte rods are 11 in.. long, 32 of them are 4 in.
Cin dlameter and 24 are / in, in diamefer. At

both “ends ~of the. contamer, retainer and baffle

were noted before the box - was sealed.
" The bifluid loop that was placed in. operat:on in

"_':March 1957 had accumulated a total of 6673 hr of . |

operation at the specified temperatures and flow

rates when operation of the loop was terminated

31

 
 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT .

  

Salts.

by the power failure des'cribed above. Freezing of

the salt ‘in the loop caused a section of the first

heater leg to rupture, This damaged tubing section
and the pump and pump motor have been replaced.
The salt circuit of the loop was refilled with a new

charge of fuel 122, but the original sodium charge’

is being used. - Operation was resumed on January
13, 1958. (The results of examinations of samples

removed from the damaged portion of this loop are

reported in Chap. 2.1 of this report.)

The loops that have been operated or are now'
operating are listed in Table 1.2.2, which also

gives operating data. A maximum pump" speed of

3000 rpm has been established on the basis of test
- ‘experlence in-order to ensure long Ilfe. At this

speed, full turbulence in the fluid is difficult to

attain, however. The Reynolds numbers are low

32

UNCLASSIFIED
~ORNL-~LR-DWG 279302

   
  

3g-in., SCHED -40
INOR-8

  

SAMPLE INSERT

Fig 1.2. '|6 Dlogrum of Forced- Circuluhon Loop for Weighi Loss Studies of INOR-8 Exposed to Circulcfing Fused

and vary widely for different flunds becouse of

dlfferences in viscosities.

In-Pile Loops
Do Bo Tl'cwger : . :
J. A. Conlin - P. A. Gnadt

- Authorization has been requested for wradlatidn

of a forced-circulation loop: m the MTR, The loop .
is to consist of a hairpin of é-m., sched-40 pipe
with addlhonal coils in the nose end near the
reactor face, a miniature pump to circulate the
~ fuel, and ‘a salt-to-air heat exchanger. Shielding,
hux:hary cooling, pump drive mechanisms, ‘and
" instrumentation will also be ‘provided.  Oper-
- ation of the loop will provide - information on.
fuel sfablllty and on the corrosion of INOR 8 under

 
 

£e

 

 

 

 

. i r? iy s n‘ o
o ‘Table 1.2.2. Forced-Circulation Loop Operations Summary as of January 31, 1958
; ST Couifio3|t|5n oo Maximum Minimum Maximum Hours of
Loop " Loop,. . Number of F'low Rate  Approximate Wail Fluid Fluid Operation ot c
Designation - Mmflol and Snze Clrculmd - {gpm) - Reynolds Number Temperature Temperature Temperature = Conditions omments
aa . - Flmd* o (°F) (°F) (°F) Given
9344-1 lnt:ldi;lelg ‘4 in.m, ‘I 123':}.“_:‘ 2“‘.0" . 3250 - 1300 1100 1210 2400 . Operation ‘c.mtmuing, salt hfis' fiozon twice
‘ 0.045.in.| :wcll_- LT e T e ‘ T ‘ .because of contral d:fficuhy and o power
T L . failure
93442 Tngnel, ,5sn.coj Coom a2l 8200 1200 1000 1100 1328 Loop failed on thaw-out attempt after power
i 0 045 in. wull R - failure; repairs undor wuy, operation to be
: o ‘ resumed .
9377-1 - |ncone| 6 in.d) " 126 f '_ 2.0 1600 1300 1100 1175 1890 Operation continving; salt froze once be-
Do . 0.045 in.wall ; S . i PO _ cause of power failure
9733-2 l'w?'jlncoml !5 in. CD. L y"l3cl o 2,0 | 3000 1300 - 1100 1210 1510 Operation cofifinuing; salt has frozen mic;
‘ L 0.045 in. woll L ' because of motor control difficulties and o
. e | \ _ power failure
9354-3 “INOR‘B L 84 “- 2.75 o 1200 1070 1150 1160 Operation continuing; salt has frozen twice
Hot leg 3, m b 4500 because of motor control difficulties and a
o sched'40 L : ) power failure
Cold leg, .n.'_ ! 5400
OD 00045 i". .
. wuli - S |
9354-1 -f_:INOR-a, 35 0D, 126 25 2000 1300 1100 1210 184 Failed on stortup; repaired and restarted,
L 0.045 in. wull ‘ - to ‘ : but flow stoppages because of plugs
finally resulted in second failure; repairs
. ‘ being made
9354-2 INOR-B, 4 in. OD 12 o ; 2.6 6500 1200 1050 1140 12 - Hastelloy pump bowl failed after 112 hr of
L 0.045 ine wall T ' : ‘operation; repairs being made
CPR - lnconel s 122 :,':‘ ' .'l o 5000** 1250 1095 1190 | 6673 Operation'confihuing; salt has frozen twice;
Sodiom T port of salt tubing domaged second time

97,700 | 1085 n3s and replaced; pump replaced and fuel re-

placed after second freeze

 

*Composition 123: NoF-BeF UF (53-46-1 mole %)
Composition 12:° NoF-LiF-f(F (11.5-46.5-42 mole %) -
" Composition 126: LiF-BeF —UF (53-46-1 mole %)
Composition 130: . LiF-BeF UF (62-37-1 mole %)
Composition 84z NoF-LanBeF (27-35-38 mote %)
Composition 122 NaF-ZrF -UF (57-42-1 molo %)

**|n heot exchangar

8S61L ‘L€ AYVYANYI 9NIAN3T QO1d3d

 
 

 

 

 

MOLTEN.SALT REACTOR PROGRAM PROGRESS REP-O.RT
simulated reactor service conditions. The operating
conditions specified are listed below:

UNCLASSIFIED
PHOTO 30631

 

Average loop power density . - 50 w/cm®

| Nose coil power density | 187 w/_c:m3
Maximum wall temperature I. 1300°F |
Maximum temperature difference ~ 150°F

(obtained by the addition of
4 kw of electric heat)

Pump speed | 1700 rpm

Flow velocity ' 2.7 fps
Propovsed test duration 2000 hr
Reynolds number 1600

Fission power 7 _ 7.8 kw

A section of the heat exchanger and the heater

~ have been fabricated, and out-of-pile tests of these

components will be initiated as soon as they can

be incorporated into an existing heat transfer test
focp.

Fig. j.2.i_7. G.ruphite Specimen Assembly for In.sertion
into INOR-S ququ_-Circulction Loop.

34

 
 

 

¥

»)

PERIOD ENDING JANUARY 31, 1958

1.3. ENGINEERING RESEARCH
H. W. Hoffman

HYDRdDYNAMIC_ STUDIES OF MSR CORE
F. E. Lynch

The molten salt reactor as presently designed
has a spherical core. Three entrance-exit systems
have been proposed for this core, namely, inlet
and outlet diametrically opposite each other with
straight-through flow, inlet and outlet concentric
with the fluid entering through the inner pipe and
exiting through the outer annulus, and inlet and
outlet concentric with the fluid entering through
the annulus and exiting through the inner pipe.
With each of these entrance-exit systems, stagna-
tion and flow separation are possible in the
spherical core region and could cause excess

heating of the core wals and thermal cycl:ng of

reactor components.

Glass models of each of the 1hree core configura- -

tiens have been fabricated from large (13.7-in.-ID)
Pyrex reaction vessels. Two of the models are

shown in Figs. 1.3.1 and 1.3.2. The straight-

 

Fig. 1.3.1. MSR Straight-Through Flow Model.’

through core has 6-in.-dia entrance and exit pipes.
The concentric system has a 4-in.-dia central pipe
within a 6-in.-dia 90-deg elbow. In the model for
annular flow studies, the central tube is flared

‘outward at the bottom to direct the flow along the

spherical surface.

Initial qualitative data have been obtained for
the concentric system with fluid entering through
the central pipe by using the phosphorescent-
particle flow-visualization technique. Both visual
observation of particle motion within a continuously
illuminated  plane and photographic recording of
instantaneous velocity profiles were employed
obtaining the data. Figure 1.3.3 shows the results
obtained visually at an inlet Reynolds modulus of
80,000 (2.58 fps in the 4-in. pipe). The inlet jet
was of sufficient strength to retain its shape (a
4-in.-dia cylinder) to the bottom of the vessel. At
this point the fluid turned and moved upward in

a‘_l/2~ to %-in.-wide annular region immediately

 UMCLASSIFIED
3 PHOTO 30531

Fig. 1.3.2, MSR Concentric Entrance-Exit Core Model
Designed for Fluid to Enter Through Inner Pipe.

35

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Y
11
I FLUID
LI

 

 

UNCLASSIFIED
ORNL-LR-DWG 27903

 

[ao— 7

 

 

FLOW UP OUTER WALL IN
LAYER ABOUT {in. TO
Yin. WIDE. "

LARGE TURBULENT EDDY
WHICH BROKE UP INTO TWO
EDDIES, WITHIN THE LARGE
EDDY, AS THE FLOW PULSATED.

  
 

. S
—_— FLUID FLOW QUT

— 6-in.~-1D GLASS PIPE
————— : .
—

-
s Y,

 

     
  
  

“— LOW VELOCITY AREA
PARTICLES SETTLE OUT

VELOCITY PROFILE OF INCOMING FLUID; AS THE
FLOW PULSATED, THE FLUID FLOWING IN FANNED
QUT INTQO THE SPHERE, AND THUS DECREASED
THE SIZE OF THE LARGE EDDY, OR THE LARGE
EDDY EXTENDED INTO THE FLUID FLOWING DOWN
THE CENTER OF THE SPHERE.

Fig. 1.3.3. Flow Patterns Observed Within an {lluminated Plane of the Concentric Entrance-Exit Core Model

with Fluid Entering Through Inner Pipe.

adjacent to’ the core wall. The volume between
these two flow regions was filled with a large
toroidal ‘eddy, with its axis of rotation lying near
the main upward flow. Because of unsteadiness
in the flow-generating system, the flow patterns

fluctuated. The main effect of these pulsations

was in the structure and size of the large eddy.
During periods of pulsation, the central core of

‘the eddy broke up into two smaller eddies sur-

rounded by the larger rotating mass. It can be

- 36

“seen that the eddy removes fluid from the exiting

stream and recirculates this liquid through the
high heat generation regions of the core. If good
mixing occurs, there could result a uniform increase
of the -exit temperature that corresponded to a
longer fluid residence time in the core. However,
there exists a strong possibility that the outgoing
stream will contain discrete fluid bodies that will
be at temperatures considerably higher than the
mixed-mean temperature. Under these conditions,

i

 
 

 

 

 

o

Hy

high-frequency thermal cychng of system com-

~ ponents could occur. -

Several typical mstantuneous proflle photogruphs
of the flow in the concentric system are shown. in

Fig. 1.3.4. Since grid photographs are not avail-

able, quantitative velocety analyses of the data
cannot be made. In all cases the time lapse be-

tween the initial excitation and the recording of

the profile was approximately 0.1 sec. In Fig.
1.3.4a the velocity profile at the exit of the inlet
pipe is shown. The straight light line above the
profile indicates the initial excitation, or zero-time
line. The zero-time line is either dlrectly marked
or indicated by a small triangle near the outer
sphere wall on all the photographs. The profile in
the outer annulus is also visible in Fig. 1.3.4a.
Figure 1.3.4b shows the profile at a position just
cbove the equator of the sphere, and both the

- downward central flow and the upward flow along
the wall may be seen. The effect of flow pulsation

may be seen clearly in Figs. 1.3.4c and 1.3.44 for

excitations at the same position but at different
_times. Figure 1.3.4¢ shows a profile obtained ot

a 45-deg angie near the sphere bottom. = This
photograph indicates a very rapid drop- off in
velocity in moving rud:olly away from the surface

ond defines clearly the extent of the low velocity -

region (high llghf-mtenslty region).

Studies will be continued with this model, as

well as with the straight-through and the annular

entrance systems. The effects on the velocity

pattern of inducing vortex motion by a generator
Iocuted at the entrance will also be defermlned

PHYS|CAL PROPERTY MEASUREMENTS
' W, D Powers e

7 Experlmentul determmuhons are. bemg que o!--,-
| rthe viscosities -and thermal conductivutles of -
 several BeF -beurmg fluoride salt’ mixtures.” The
o composmons to be studied have been: prepared ]
" and preliminary ‘measurements are being made on
" _one mixture. The tox;c:ty of the beryllium salts .

"‘_‘dactuted that oll meusurements be made -in |nerf—‘_,

- cup (capillary) viscometer will be used. " This:

:f‘mstrument ‘consists’ of a 1 m.—long ccp:llary tube . .
- _through “which the fluid- dralns from a’reservoir
~ with a capacity of about 6 cm®. The time required

for the reservoir to empty ,through the tube is
directly proportional to the kinematic viscosity.

The cups to be used have been calibrated with:

"PERIOD ENDING JANUARY 31, 1958

glycerine-water solutions whose viscosities have

been accurately obtained wnh Cannon-Fenske-

Ostwald viscometers.

The thermal conductivities of several Ber-
confmmng fluorlde salt mixtures will be measured
with the -use of a varicble-gap device. In this
apparatus the molten salt is contained in the gop
between two parallel plates, the upper plate being
movable. A series of measurements are made for
various salt thicknesses. This technique permits
the separation of interfacial and metal thermal
resistances so that the thermal conductivity of the
salt alone is obtained.

MOLTEN SALT HEAT TRANSFER STUDIES
H. W. Hoffman

Heat transfer studies of molten scltsl'z have
indicated that the general correlations for heat
transfer of ordinary fluids (0.5 < NP < 100) also
apply to the salts. -However, these investigations

_have also shown that for some metal-salt combina-
tions marked reductions in system heat transfer

occur because of interfacial film formation and
nonwetting.. Therefore, data.on these phenomena

are needed for the desngn of crmccl heat trunsfer
“components.’ ' :

The apparent heat trunsfer coeffncnenf experi-

'mentully determined for flow through a single
_circular tube has been found to be a sensitive
_indication of the presence of nonwetting or inter-
~facial film formation in molten salt systems. The
_experimental apparatus developed for such heat
- transfer coefficient determinations is shown in.

Fig. '1.3.5, and @ schematic diagram of the system

" _is shown in Fig. 1.3.6. It is proposed to use this
"'-.;’apparufus to study the salt mixture LiF- Ber-UF4
" {53-46-1 _mole %) ‘in- both . Inconel ‘and INOR-8
~ systems. - The molten salt, contained in the two
‘tanks, is ¢ycled through an electrical-resistance-
_heated test section by pressurizing one of the
“tanks with the helium ‘blanket gas while venting -
-the other tank. ‘When the salt has been transferreds

S mto fhe vented tonk 1he pressures are uutomahcu“y
" gas dry boxes.. For the vuscosnfy study, the efflux o T -

 

IH. ‘W. Hoffindn, 'Turbflul‘erirt Fd}cea-Cbnvectzbn H’eat 7

; Transler in Circular Tubes - Contammg Molten Sodmm
Hydroxide, ORNL-1370 (Oct. 20 1952),

2H W, Hoffmon and S. I, Cohen, Fused Salt Heat
Transfer Part lI; Forced-Convection Heat Transfer
in Circular Tubes Containing the Salt Mixture NaNO ,-
NaNOB-KNO ORNL.-2433 (to be published).

37

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

 

 

@ W

 

c) - )y

 

(e) :

Fig. 1.3.4. Insfantuheous-VeIocity-Pffi?ile 'Phofogrdph# in 'Concenfr'i'c.;E_iiirdncé-Exit_ Core _Modd with' Fluid.
Entering Through lnné; Ffipe.‘ : o ' S

38

    

Y

O

 
af

¥

»

 

 

   

 

b

 

 

 

Fig. -1.3.5. Experimentol sttem for

Mole %).

 

Defermifiing Heat T.l"unsfer

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED
PHOTO 28430

 

Coefficients for LiF-Ber-UF4 (53-46-1

39

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS.REPORT

reversed, and the flu:d is coused to flow in the
opposite direction through the test section. The

fluid flow rate is determined by observmg the

deflection of the steel beam that supports one of
the tanks as the fluid flows into or out of the tank.

Test section inlet and outlet mixed-mean tempera- )
tures, as well as tube outside surface temperatures,

are recorded. - Because of the toxicity of the
beryllium salt,

traps or absorbers on the system gas vent lines.

it will be necessary to provide

-Sufficient salt has been prepared to charge this
system, and experimental measurements will be
_made in the near future.

Since film formation, if such should occur, may

“be a relatively slow - process, an apparatus is
_bemg designed which will allow long-time, continu-

ous exposure of test section tubes to flowing salt.

_ The tubés will then be removed from this system

and welded into the heat m:msf_er coefficient

apparatus for study. .

" UNCLASSIFIED
ORNL—LR—DWG 27904

 

CYLINDRICAL STEEL BEAM _

 

VENT 4 * He SUPPLY ¢

 

 

:En —~—BEAM DEFLECTION
INDICATOR

VENT 1

 

 

:

Lb—SOLENOID- VALVE ” SOLENOID- VALVE—|J|]

E LECT RODES

— FLEXIBLE HOSE

O

 

SWIVEL JOINT

 

 

 

 

. CONCRETE 3
- Po T :'->

 

 

—

SUMP TANK

 

 

 

 

NS TEST
~ SECTION

\Muxms CHAMBERS

FLEXIBLE BELLOWS

 

 

  

 

 

bk

. ————

L/ "WEIGH TANK

 

 

 

 

 

 

 

 

 

Fig. 1.3.6.
BeF,-UF, (53-46-1 Mole %)..

'Séfiemati;: Diagram of E‘xperimehtu.l;System for Detrerminingyl'leui Transfer Coefficients for LiF-

 
 

PERIOD ENDING JANUARY 31, 1958
1.4. - ADVANCED REACTOR STUDIES

A MOLTEN SALT NATURAL-CONVECTION | B UNCLASSIFIED

_ REACTOR = ORNL-LR-DWG 27943
F. E. Romie B. W. Kinyon EXPANSION DOME

    
    

One of the problems of a circulating-liquid-fuel
reactor is the provision of reliable iong-lived fuel-
circulation pumps. This problem would be elimi-
nated if the liquid fuel circulated by natural
convection through the reactor core, through vertical
convection risers, and through the primary heat
exchangers. A molten salt fuel is well suited to
such a reactor because high temperatures could be
attained without pressurization, and therefore a
gas-turbine cycle or one of several steam cycles
that operate efficiently under high-temperature
conditions could be utilized. The temperatures
would not be so high as to be inconsistent with
the desired long corrosion life. The advantages
of eliminating the fuel-circulation pump and the
attendant problems of maintenance and replacement
are obtained at the cost of the. increased fuel
volume required for a system in which the pressure
losses must be very low. There are, however,
applications for a reactor system in which the
premium placed on reliability and ease of mainte-
nance could make the natural-convection. system
attractive. The results of a preliminary design
study of such a system are described bruefly here,
and a detailed report is being published.”

A schematic diagram . of the 60-Mw (thermal)

‘molten salt natural-convection reactor system that _

. was unalyzed is shown in Fig. 1.4.1. The. temper-. L .
~ ature of ‘the fuel suh enterlng the: exchanger was_f*-' R
 ~specified to. ‘be - T225°F “a temperature that s ..
. ‘consistent w:th long ‘corrosion life, ‘and the exlf'f;;-:;}.'g Ny
o fuel-salt- 1emperature was varied from 975 t0'1025°F.. Ly
"' The fuel salt used in'this study was LiF-BeF, UF Y A e
- (62-37-1 mole %, mixture 130). . Two med:ums a‘ R

- molten - salt (NcF-LlF BeF .27-35- 38 mole %o,

- mixture 84) and helium, were cons:dered as cooldnts S
_for the. prlmary exchanger. - The exchanger entrance .
~and ‘exit temperatures: for the ‘molten salt coolant -

- were fixed at 875 and. 1025°F, respectively. These'j-’: T
‘ jtemperutures are consistent. with the generation of
~850-psia, 900°F steam in a sysi‘em in which heat- . . -0

HEAT EXCHANGER

COOLANT 1

DOWNCOMER

"

 

- s tronsferred from the fuel salt to a- coolunt sald -
F. E. Romie and B. W. Kinyon, A Molten-Salt Natural~ : :
u Convection Reactor System, ORNL CF 58-2-46 (t;‘ ‘ie Fig. 1.4.1. Schematic Diagram of o Molten Salt
published). " Natural-Convection Reactor.

41

 

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

to sodium to steam. A similar system has been
descnbed elsewhere.? For a 60-Mw thermal output,
" the 850-psia, 900°F steam would give a generator
output of about 22 Mw, with a thermal effncnency
of 37%.

Two sets of terminal temperatures were specified
for the helium coolant. The first set, 850 to
. 1025°F, was selected for generation of 850-psia,
900°F steam in a system in which helium would
be the only intermediate medium between the fuel
salt and steam. The second set of helium terminal
temperatures, 676 to 1100°F, was selected for a
helium gas-turbine cycle with an estimated output
of roughly 18 Mw. '

For a specified flow rate of the fuel salt, temper-
ature change of the fuel salt, and pressure drop
across the exchanger, there is one combination of
riser diameter and height of exchanger above the
core for which the salt contained in the risers is a
minimum. - These optimum combinations were used
in heat exchanger calculations carried out for
exchanger heights above the reactor of 5, 10, and

20 ft.

Salt-Cooled Heat Exchongers ,

Design values for the salt-cooled heat exchanger
are summarized in Table 1.4.1. The designs are
based on 0.634-in.-ID, 0.059-in.-wall tubes. Smaller-
diameter tubes would give a lower total fuel-salt
volume, but the number of tubes required would be

 

2B. W. Kinyon and F, E. Romie, Two Power Genera-
tion Systems for a Molten Fluoride Reactor, paper to be
presented at the 4th Nuclear Engineering and Science
Conference of the 1958 Nuclear Congress, Chicago, lll.,
March 17-21.

larger. For example, 0.42-in.-ID tubes in an ex-

~‘changer 10 ft above the reactor would involve a

total external fuel-salt volume of 123 #3, but it
would require 7200 tubes .

The design valves presented in Table 1 4.1 are
based on a fuel-salt temperature change of 225°F.

Varying the temperature change from 200 to 250°F

does not change the total fuel-salt volume appreci-
ably. '

Helium-Cooled Heat Exéhufigers for Steam Cycle

Reference design values for a helium-cooled heat
exchanger in which the helium terminal temperature
would be suitable for the generation of 900°F
steam are presented in Table 1.4.2. It was assumed

Table 1.4.2. Design Values for a Hélium-Coolé_d
Heat Exchanger for a Steam Cycle

Helium temperature range: 850 to 1025°F

 

Helium pressure level: 100 psia

Height of heat exchanger above 10
reactor, ft

Number of tubes 3240

Length of each tube, ft 4 1.5

Distance from front to rear of tube 0.77
array, ft

Transverse dimension of heat 78
exchanger, ft

Generator output used to pump 2

helium through exchanger, %

Total fuel-salt volume outside core, ft3 172

 

Table 1.4.1. Design Values for a Salt-Cooled Heat Exchanger for a Natural-Convection Molten Salt Reactor

Coolant salt temperature range: 875 to 1025°F

 

Height of heat exchanger above reactor, ft
Number of tubes
Diameter of tube bundle, ft
Length of each tube, ft
Fuel-salt volume in risers, ft3
Fuel-salt volume in headers and tubes, ft°

Total fuel-salt volume outside reactor core, ft3

5 w20
4570 3150 2200
5.1 | 4.2 3.5
7.8 ns - 16

. 55 68 89
97 91 | 86
152 159 s

 

42

 
o

i

 

4@

that copper fins would be used on the helium-cooled
heat exchanger tubes. The use of lower-conduc-
tivity fins would lead to a considerably increased
fuel inventory. The dimensions given for the heat
exchanger are those for a three-helium-pass cross-
flow exchanger. - Doubling the helium pressure
level to 200 psia while maintaining a pumping
expenditure of 2% of generator output would not
change the total salt volume appreciably but would
give a more compact heat exchanger. A similar
result would be obtained if at 100 psia the pumping
expenditure were doubled. , -

A cross section of a possible helium-cooled heat
exchanger configuration is shown in Fig. 1.4.2.
The configuration shown permits attachment of the
riser pipes to the salt headers, with sufficient
flexibility to minimize thermal stresses, and pro-
vides a cylindrical container for the pressurized
helium. . Use of two such cylinders to contain the
reference-design heat exchanger would give two

PERIOD ENDING JANUARY 31, 1958

heat exchdnflgers,'eoch 19.5 # lon.g.r If the helium
pressure were increased to 200 psia, the length

‘of each H_ed_t exchanger would be reduced to 12 ft.

Helium-Cooled Heat Exchanger fbr
Gas-Turbine Cycle

The temperature of the helium returned to the
heat exchanger from the gas-turbine-system regen-
erator is estimated to be 676°F, which is 174°F
below the fusion temperature of the fuel salt.
Therefore a counterflow heat exchanger with
fongitudinally finned tubes was selected for the
gas-turbine cycle. The temperature of the interface

‘between the fuel salt and the tube wall can be

maintained at, or above, any desired value by
adjustment of the gas-side thermal resistance per

unit tube length. The design values given ‘in
~ Table 1.4.3 are based on a minimum interface

temperature of 900°F. For a longitudinally _finned_

UNCLASSIFIED
ORNL-LR- DWG 27948

    
 

_-PRESSURE SHELL

Fig. 1.4.2, Schematic D.iugrum of a Possible -FueI-Suit—to-Heliun'.l Heat Exchanger with Three Gas Passes.

43

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Table 1.4.3. Design Values for a Helium-Cooled |
Heat Exchanger for a Gas-Turbine Cycle

Helium temperature range: 676 to 1100°F

 

Height of heat exchanger above 10
reactor, ft

 Number of tubes 3300 -
Length per tube, fi ' 1
Total fuel-salt volume outside | 160

core, ft3- :

 

tube the required gas-side thermal resistance
variation along the tube length can be obtained by
an axial variation of the fin height. No data were
available on the heat transfer and flow friction
characteristics of longitudinally finned tubes in a

‘tube bundle, and the exchanger designs are thus’

not complete. ‘However, the number of tubes,
length per tube, and fuel-salt holdup volume can
be estimated with good accuracy if it is assumed,
as seems likely, that the same gas-side thermal
resistance is realizable with uncut longitudinally
fined tubes as with circumferentially finned tubes.

Comparison of the Yarious Cooling Systems

The fuel-salt holdup volume differs less than
10% for the three cooling systems considered and
is therefore not a determining factor in the selection

of the coolant medium for the reactor. In general,

the helium-cooled heat exchangers have much
larger over-all dimensions than the salt-cooled
heat exchangers. The increased bulk is caused
by the larger spacing required by the finned tubes
and also by the large volume required by the
helium headers. An increase in helium pressure
would decrease the helium header volume and also
should decrease the fuel-salt header volume. An
upper limit on the helium pressure would probably
be set by consideration of the effects of a tube
rupture in the fuel-salt system. It is of interest
to note that, in other studies of gas cooling in
which such considerations were apparently not
limiting, gas pressures as high as 1000 psia have
been recommended.

Comj:nrison of Natural-Convection System with
Forced-Convection System

The design of a forced-convection reactor sys-
tem? led to an estimate of 0.56 #3 of fuel salt

44

external to a molten salt reactor per thermal

“megawatt. For the free-convection system using

a 160-ft3 external salt volume in the generation
of 60 Mw, the corresponding specific volume is
2.67 3 /Mw. Calculations based on these numbers
and initial, clean, critical-mass data for an un-

‘reflected molten salt reactor indicate that the fuel

inventory for both the free- and forced-convection
systems is ot a minimum with a core diameter of

about 8 ft for a thermal output of 60 Mw. For an

8-ft-dia core the specific powers ar'e'895rand
1275 kw of heat per kilogram of U235 for the free-
and forced-convection systems, respectively. The

- free-convection system is thus estimated to require
‘a fuel inventory about 42% greater than that of the

forced-convection system. With higher thermal
outputs the specific powers would be larger, but

~ the specific power for the forced-convection reactor
increases more rapidly with increasing power than

does that of the free-convection reactor.

HIGH-FLUX REACTORS
W. K. Ergen

A systematic study® was started to determine
the influence of various factors on the power

‘required to obtain a given flux. At the beginning

of the study, a reactor idealized in ‘the manner
described below was considered.?

The fuel is concentrated in a spherical shell
embedded in an infinite moderator; that is, the
moderator occupies the space inside as well as
outside the shell. The shell is “infinitely thin'’
but ‘““black’’ for thermal neutrons; that is, the
thickness of the shell is small compared with its
radius but large compared with the diffusion length
in the fuel. The shell emits nf fission neutrons
for each thermal neutron absorbed. Absorptions
at epithermal energies, both in the fuel and the
moderator, are neglected. In this model, the

fission neutrons emitted from the shell slow down

according to the age kernel® and then diffuse

 

~ Chain Reactions:

3w, K. Ergen, Preliminary Design Data for a Circulating
Fluoride~Fuel High«Flux Reactor, ORNL CF-56-6-9,

7 Revision No. 2 (Jan. 28, 1958).

4W. K. Ergen, Flux Distribution in a Reactor Con-
sisting of a Spherical Shell of Fuel in a Infinite
Moderator, ORNL. CF-57-12-100 (Dec. 24, 1957).

sA. M. Weinberg and L. C. Noderer, Theory of Neutron
Volume 1, Digus:'on and Slowing
?ggm” of Neutrons, ORNL CF-51-5-98, p 111-38 (qu 15,

 
 

ny

 

 

n

[}

v

according to diffusion theory, with the boundary
condition created by the ‘‘black’’ shell.

For this model, it is possible to compute the
number, J, of neutrons which return, at thermal

energy, to the shell per fission neutron emitted.

The neutrons returning to the shell give Jnf fission
neutrons, and if the reactor is to be critical

Inf = 1.

From this equation, the required 75f has been
computed‘ for D20, Be, BeO, and C moderators
and for various values of shell radius. The results
are plotted in Figs. 1.4.3 through 1.4.6. The

abscissa of each plot gives the shell radius, r, in

cm, aond also in dimensionless units p = r/\NT,
where /T is the slowing-down length.

In the sphere enclosed by the shell (the island,

or internal thermal column, or flux trap) the neutrons

 

5w, K. Ergen, Fluxes Obtainable in a Flux-Trap
Reactor, ORNL CF-58-1-4 (to be published).

PERIOD ENDING JANUARY 31, 1958

reach thermal energies at some distance from the
shell, and, in the process of diffusing toward the
shell, these neutrons set up a flux gradient, which
results in high fluxes at the center. These center
fluxes are also plotted in Figs. 1.4.3 through 1.4.6.
It may be seen that for Be and BeO a flux of 10-3
neutrons/cm? per fission neutron emitted is ob-
tained, which corresponds to about 140 Mw for
10'¢ neutrons/cm?.sec. For D,0 and C the flux
per fission neutron is smaller, and hence the power
required for 1016 neutrons/cm?.sec is higher. This
is primarily due to the large diffusion constants
of D,O and C, which make it possible for the
neutrons to reach the shell without setting up a
very large flux gradient.

The curves are, of course, applicable only to the
idealized model. Present investigations cover the
modifications caused by more realistic assumptions,
such as holes in the shell, finite thickness of the
shell, moderation in the shell, and flux depression
in the flux trap as a result of insertion of absorbing
samples.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNELASSIFIED
ORNL—LR—DWG 27505
ricm)
12 19 16 i8 20 22 24 26 28 30 32 34 36 38 40 42 44
(X 1074 | I | B T ] | 1 11 I ] T T3
g, 2-8
o
2
2 8 2.6
o
§
37 2.4
c =
s -
T8 T T —— 22 &
g 1 1o L &
€0 N0 | T LT~ : g
s - ~ : o e - . &
@ _ ~ B : - . \ . - =
- E4 - T : - T _"‘\ — 18 5
< < b Tao - S g
=5 - W . .
= h S R — ¢
o " Ssel g Ee :
2 -.:.‘.',-,__:7__,;. 1 ’ ‘ = ‘
Ee = i : S ) 4
wo ) . o e . s
o o " - - —— 1 i
Y : 1.2
o 1.0

D90 42 0 44 8 18 20

2224 26, 28 030 32 .34 36 38 40
;- pIDIMENSIONLESS) L ' '

Fig. 1.4.3. Central Flux and Multiplication Factor as Functions of Shell Radius of 1dealized Flux Trap Reactor

witha D 207 Moderator.

45

 
 

 

B MOLTEN.SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ORNL~LR —DWG 27906
r{cm) .
2 {0 2 4 16 i8 . 20 22 24 26 28 30 32 34 36 38 30"
X107 ] l I I T I I I [ [ I | I ! I -
9 7 ~~. ' 28
: \ " N -8
- \ / \@c
2 N '
_‘_-.E: 8 \\ / ™ - 26
@ . \ ' )
g N\ \
£ N\
a7 - 2.4 s
c
g “\ - &
£ e > AN 22 F
g N \ g
o \\ \ =
E b, " g
{ 5 ‘\\ . \ 2-0 -
g “wb ) \ . 8
.E.‘ S : o E
=2 "-..-
£ 4 = P~ 18 F
OQ- ) - "‘-l-.--, .,___.----- i g
§ 3 = 16
i .
|
a .
Eo2 1.4
E
wl
o
1 12
0 i0
1.0 2 14 16 18 2.0 2.2 24 26 28 3.0 32 3.4 36 38 4.0
p (DIMENSIONLESS)

Fig. 1.4.4. Central Flux and Multiplication Factor as Functions of Shell Radius of Ideuiized.Fqu. Trap Reactor
‘with a Be Moderator,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
ORNL—LR-DWG 27907
ricm)
12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 :

—4

(x107%) r I T | | | T 1 | | [ | | I ] 3°

9 s - 28
/ \ ¢c
3 pd ,
£8 26
§ \ 4
5 \/
.; 7 \\ - 2.4 “
g S ’ =
5 €
"
26 S oS 2.2
- 2 Y =
: N S
o~ \\ =
E ~ .o
{5 < 20 E
E ‘\~\,qf g -
\\

- 3 ~. _ g
2a S ~ 48 3
efli ..-"-.,_-‘- ) i =
§ 3 -—--- = m e _.___.__----“--- 1_6
It " s epam
-l
& A
Ee 14
=z .

Lt
Q
4 12
.0 1.0
10 2 14 16 1.8 20 22 24 26 2.8 30 32 34 36 38 4.0
' p{DIMENSIONLESS) - o ' ' o

Fig. 1.4,5. Central Flux and Multiplica

- with a BeO Moderator.

46

tion .Factor' as Functions of Shell Radius of |dealized Flux Trap Rec.cfor

 
 

PERIOD ENDING JANUARY 31, 1958

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
ORNL—LR—DWG 27908
':: rlem)
16 20 24 28 . 32 36 40 44 48 52 56 60
—g . - 30
(xt0 7 F] l J ] ] | | I | 1 i
- 9 2.8
3
€ B 26
L4
s \
2.7 LS 2.4
< \\ -
o
B &=
£ \, <
5 8 2.2 ©
. AN 5
NE . &
O LY F
=5 N7 205
g \*s '<-[
=2 ~\~ 9
< e &
:&, 4 L= “%\_ 1.8 5
. / . ) "-...,__..‘~ \& g
>3< . . —~ ~~a.l_ P — 6
[ '--..---..--..___J}‘
;I‘ T .--Q'-—---
E 2 P — 1.4
=z T iy
i
O
4 1.2
o . 1.0
i 12 1.4 .6 1.8 20 22 2.4 26 28 30 32 3.4 36 38 40

p (DIMENSIONLESS}

Fig. 1.4.6. Central Fiux and Multiplication Factor as Functions of Shell Radius of {dealized Fiux Trap Reactor
with a C Moderator.

 

»

 

i

!

! .

11 .

| - | S 47
j. . .- ‘ . .

 
 

 

"

Part 2

MATERIALS STUDIES

 
 

 

.

 

 

 
 

 

.

L]

W. D. Manly

D YNAMIC CORROSION STUDIES

Jl Ho Devan
J. R. DiStefano. R. S. Crouse

Corrosion experiments are under way for which
thermal-convection loops and forced-circulation
loops were fabricated of Inconel (nominal compo-
sition: 15 wt % Cr, 7 wt % Fe, bal Ni) and of
INOR-8 (nominal composition: 17 wt % Mo,
6 wt % Fe, 6 wt % Cr, bal Ni). As discussed
previously, the test program is being conducted in
three distinct phases,! In the first phase, relative
corrosion properties of 12 fluoride sait mixtures
are being determined in thermal-convection loops
operated for 1000 hr. During the quarter, tests of
Inconel loops were completed for all the salts
except salt 131 (LiF-BeF,-UF ,, 60-36-4 mole %);
four salts were tested in INOR-8 loops, which are
now being examined. As part of phase 2 of the
program, three salts are being circulated in Inconel
loops and five in INOR-8 loops. The tests will be
run for extended periods at two temperature levels,
1250 and 1350°F. In the third and final phase of
the test program, salts are being tested in forced-
circulation loops under simulated reactor operating
conditions. Four salts are presently being tested
in Inconel loops and three in INOR-8 loops, as
described in Chap. 1.2 of this report. The results
of postoperative examinations of these loops will
be described in this chapter as they are completed.

Ten of i'he Inconel loops operated in phase 1 of

the progrom have been examined metallographically,

;'After successful .operation for ‘1000 ‘hr, the salt
mixture - in each’ loop was allowed ‘to freeze in =

~ place. -Samples were then cut from each loop at -

- the locations shown in Fig. 211, and the salt. _

. 'contamed in"each sample was ‘melted ouf under a

= helivm utmosphere. The results of metc:llographlc,j

o ;excmsnatlori of the samples are presented in Table_';'rij

- 21 land are dlscussed below. ' ' '

Thonum-Bearmg Sulis in 'I'hermul-Convechon
: : Loops R L

The loops whlch cnrculated thorsum-bearmg saltsv'_rr'

comprlse three pairs on- the ‘basis -of slmllarmes

 

1J.' H. DeVan, J. R. DiStefano, and R. S. Crduse, MSR-

Quar. Prog. Rep. Oct. 31, 1957, ORNL-2431, p 23.

A. Taboada

in the fuels they circulated, Loops 1169 and
1177, the first pair listed in Table 2.1.1, circu-
lated salt 128 (LiF-ThF,, 71-29 mole %). Loop
1169 showed less than 1 mil of attack throughout
(Fig. 2.1.2), while loop 1177 was, in general,
attacked to a depth of less than 1 mil (Fig. 2.1.3)
but had scattered pits to a depth of 1.5 mils in
some areas.

UNCLASSIFIED
ORNL-LR-DWG 23451

~ | METALLOGRAPHIC
SAMPLES

HOT—LEG

SIX 6-in. CLAM- CHEMISTRY SAMPLE

SHELL HEATERS

    

COLD—LEG

CHEMISTRY SAMPLE

~ | METALLOGRAPHIC
SAMPLES

 

 

 

 

Fig. 2.1.1. Diagram of a Standard Inconel Thermal-
Convection Loop Showing Location of Metallographic

- ‘Samples.

The sécond pair of loops, 1173 and 1176, was

.. - operated to evaluate salt mixtures 124 (NaF-BeF ,-
- ThF,, 58-35-7 mole %) and 127 (LlF-BeF -ThF4
© 58-35-7 mole %). . As may be seen in. Table 2.1.1,

loop 1173, in which sait 124 was circulated, was

- - attacked substanhally more than foop 1176, which
. circulated” salt 127, * This result is in conflict,
- -however, with the - corrosuon properties normally
 exhibited by NaF-BeF ‘and LlF-BeF2 mixtures;
~in previous tests the LlF-BeF mixtures. tended
:to produce more initial corroslon in fnconel sys-

tems than NaF-BeF2 mixtures produced, Thus the
test results for loop 1173 are questionable, and o
repeat test under similar conditions has been

51

 
 

 

 

 

 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Tuble 2.'l 1. Results of Memllogrnphlc Examinations of Inconel Thermcll-COrwecflon Loops '

Operuted in Phuse 1of Ccrrosion Test Progrom

 

o Sqlt

- Mefo_!logrdphig: Results

 

H p_f;:L"e g

 

Loop NS; Designation | Salt-:vC“o_m-posifion PR ,
‘ _ _ o - Attack - - Cold-Leg Appearance
(mils) -
_ o | > The'rium-Bea;'ing So't_s .
nee 128 LiF-ThF,, 71-29 ;5;'1; % <1 Light surface roughness
177 128 LiF-ThF, 7129 mole % s Attack 1o o depth of | mil
1’1_73_' - 124 NaF-BeF,-ThF 4,'53-"35-7 mole % 4 - Attack to a depth of 1 mil
| 1176 ,lé7 | LiF-Ber-ThF4,V"758-35-,7 n;lole % <1 | Surfuce roughness and surfoce pits
1174 . 125 NaF- BeF'2~U F4-th4; 53-46-0.5-0.5 md_le'%; 2 . Surface roughness and surface pits
1163 .' i23 . Na F-Be‘Fé-UF4,53f4e;'|mole % <l - "_ ‘:St'nr:leavi':e rougflness
Urunilum-_Beor-ing Salts and Coolant Salts ' - ’
1161+ 122 NaF-ZrF -UF , 57-42-1 mole % <1 - Attack to a depth of <1 mil
17 129 NaF-ZrF -UF ;, 55.3-40.7-4 mole % 2 .' " Very heavy sufface roughness
: s and pits.
nn '_ 126 LiF-Ber-UF4, 53-46-1 mole % 2 Voids to a depth of 1'mit .
n2 84 LiF-NaF-BeF,, 35-27-38 mole % 2 Attack to a depth of <1 mil
ws 12 LiF-NaF-KF, 46;5-1-1.5-42 mole % <1 Attack to a depth of <1 mil

 

~ Table 2.1.1.

*Operated prior to the period of this report.

initiated, The fuel salts used in loops 1174 and

1163 were identical except that the fuel in loop
1174 had 0.5 mole % ThF, substituted for one-

half the UF, in salt 123. There was attack to a
depth of 2 mils in loop 1174 (Fig. 2.1.4), whereas

loop 1163 was attacked to a depth of o;r'll'y'"'l_ mil,

' _-Ueuhfh‘m¥Bearing Salts and Coolant Salt-s in
" Thermal-Convection Loops

Four loops (1170, 1171, 1172, and 1175) that

curculoted uranium-bearing or coolant saits were
also exumlned ‘and the results are presented in

1161, which was operated previously with a salt
similar to thcf circulated in loop 1170 except that

52

Loop 1170, whmh.,c_ucu_lated a
zirconium-base mixture containing 4 mole % UF,,
‘salt 129, was operated for comparison with loop

the UF concentrcflon was a factor of 4 less. The

“attack in loop 1161 was to a depth of less than 1

mil, compared with 2 mils in loop 1170, These
results emphasize the importance of the UF , con-

centration on corrosion by fluoride salt mixtures.

" The operafidfi'of loop - 171 gave the first data-.

for evaluation of an LiF-BeF -base fuel system
and revealed. shghtly more attack than has been

found - for either ‘the NaF- BeF -UF or the NaF-

ZrF -UF ; fuel systems. Whlle the cttack in this

Ioop was, in general, less than 2 mils deep, it was

~ quite concentrated, as may be seen in Fig. 2.1.5.

_Thewcichl_qfie_n of the coolant salts 12 and 84 at
@ maximum temperature of 1125°F resulted in attack
- similar to that found in the loops that circulated

fuel-bearing mixtures, Attack in loop 1172, which

"

n

 
 

 

 

20!
g0z
53

- f !
gl saHont  §

007
.008
010
011

Loi2
013
o8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o)
003
o
us
-
o
Z
006
007
008
o10
018
o2
O3
014

 

 

 

 

 

 

PERIOD ENDING JANUARY 31, 1958
Specimen Taken from Hot Leg of Inconel
Specimen Taken frblfi Hot Leg of Inconel

-Convection Loop 1174 at Point of Maximum
Convection Loop 1171 at Point of Maximum

 

~ <
s ¢
2 , 3
-

. g ) o

- E L & _

TIF -
-m-_—”mluv”- umum [« X
L = o w 23
= 3 =

 

 

 

 

L0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Specimen Taken from Hot Leg of Inconel
Specimen Taken from Hot Leg of Incone!

-Convection Loop 1169 ot Point of Muxi_mum
-Convection Loop 1177 at Point of Maximum

| o
R

& &

§ ;

: {

-

o o

- ”_-.. - -..w

, “ - < a.

~ E - £

NBE N g

.m.mwr , .m-.mo.

L ¢ o w 29

= -

 

 

 

 
 

 

‘MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

“circulated salt 84, was more extensive with re-
spect both to amount and to depth than the cm'cck
in loop ]175 whlch circulated salt 12,

Results of Examination of Samples Removed
from Forced-Circulation Loop CPR

As stated in Chap. 1.2, samples were taken for

metallographic examination from the forced-circu-

lation bifluid loop designated CPR., The samples
were taken from a portion of the hot leg which
ruptured when the salt froze because of a power
failure after more than 6500 hr of operation, A
sample taken from the discharge end of the heated

leg showed no measurable corrosion (Fig. 2.1.6),

but samples taken from near the point at which
the failure occurred, that is, near the iniet to the
heated ‘leg, revealed attack to a depth of 2 mils
(Flg. 2,1.7). The deeper attack at this point was
- probably the result of fuel-contamination near the
“point of failure, The loop was successfully re-
paired and is now back in operation,

GENERAL CORROSION STUDIES

E. E. Hoffman
W H. Cook D. H. Jansen

Efiect of Carburization on Reucfor
_S..fl_'uctural Materials

Studies are under way for determining whether
the mechanical properties of INOR-8, Inconel,

 

 

 

 

Fig. 2.1.6. Section Taken from Inconel Forc'ed-CIrf'.'u'—
Iuhon lLoop CPR at Discharge End of Heuted Leg.
250X Reduced 22%.

54

 

and potential reactor structurai alloys would be
detnmentally altered by carburlzahon. Tests

have indicated that fl1e sodlum-graphlte system is
a rapid and effective carburizing medium for stain-
less steels, Hastelloy B, and Incone! upon ex-
posure at 1500°F for 100 hr. Carburization tests
of Hastelloy B are of particular mterest because
the composition of Hastelloy B is similar to that
of INOR-8, both alloys being of the nickel-molyb-
denum-chromium-iron system. A summary is pre-
sented in Table 2,1.2 of the results of static tests

T Y
INCHES
i 3

B

EEREBEE

<
o

A
= »

o |

.28

 

 

   

Fig. 2.1.7. Sechon Taken from Inconel Forced-Circu-

lation Loop CPR at Inlet to Heated Leg. 250X, Re-

duced 22%.

Table 2.1.2, Summary of Results of 500-hr Static
Carburization Tests of Hasfellqy B Specimens__

Nominal specimen dimensions: 0,035 x 0.4 x 0.7 in..

 

 

. ‘ Test Depth of
Carburizing IR
Medi : Temperature Carburization
iu : ,
earm (°F) (mils)
Sodium + 25 wi % 1500 29w
graphite* - 1200 4
Fuel 30 + 25 wt % 1500 3
graphite* .‘ 1200 T 0

 

*Redctor grade.

**Carbon completely penetrafed the 0 035-m.-th|ck
specimen; the value given is depth of penetration of the
Hastelloy B copsule in which the specumen was exposed.

) )

 
 

 

 

 

Qb o

'of Hastelloy B in sodlum and in NaF ZrF4-UF
- (50-46-4 ‘mole %, fuel 30) contammg 25 wt %
. grophlte. The effects of temperatire and medium
~on the carburlzohon of this alloy are clearly shown,
The ‘specimen - that “was exposed at 1500°F to
fuel containing graphlte is’ shown in Fig. 2.1.8,
The carbide-like - ‘material on- the edge of the
‘spec:imen was not uniform; it varied from zero to
a maximum depth of I mils. - :
- The carburlzotlon of Inconel v was mvestagated in
100-hr- seesaw-furnace tests in which the hot-
- and cold-zone temperatures were 1500 and ]250°F '
respechvely._ "The ‘test . copsules were. mode from
l-in., sched- 40 lnconel pipe, in which was sus-
*  pended, near the cenfrul axis, a C-18 (commerccal-
- ‘grade) graphlte rod ¥ /g in. in diameter and 18 in,
long. The capsules were filled with fuel 30 A
_ copsule wnthout the grap‘nfle rod was tested under
- identical ‘conditions. as: a centrol.
-of speclmens token from the hot and cold zones

 

Examlnatlons‘ ’

PERIOD ENDING JANUARY 31, 1958

:of the pipes after the tests did not reveal any

positive - mdlcahons of carburization. The exami-
nations lncluded hardness measurements, metallo-

'grophlc examination, and chemical analyses.

Additional data were ‘obtained from two Inconel
thermal-convection loops with graphite inserts in
which fuel 30 was circulated,” These loops (Nos.
30 and 39) were constructed as shown in Fig.
2.1.9.  The graphite insert in loop 30 was the
commercial material designated C-18, which had an
average bulk density of 1,60, and the insert in

loop 39 was the commercial material designated

CCN, which had an average bulk density of 1.90,

‘The hot legs of both loops were at 1300°F, The

cold leg of loop 30 was at 1040°F and that of

" loop 39 was-at 1070°F, Operation of loop 30 was
‘terminated at 695 hr, rather than the scheduled
1000 hr, because of a power failure; loop 39 oper-
ated the full scheduled 1000 hr.

 

 

 

 

 

 

 

Flg. 2 l 8. Hos!e!lOy B After Exposure to Fuel 30 Conimmng 25wt %. Graphlfe at 1500°F for 500 hr, Etchant:

: chromrc cnd hydrochlonc acids. 750)(

55

 

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

FILL POTS

 

 

. GRAPHITE
© INSERT
ASSEMBLED

" HOT LEG

' GRAPHITE INSERTS °,

COLD LEG

 

 

 

 

Fig. 2.1.9. Inconel :Tlierrl-\ul-Convgc'fl“é_n : Loqp with =

Graphite Insert.

Mefai_logrdphic examinations were made of speci-
mens taken from positions indicated by circled

numbers in Fig. 2.1.10. The regions that were
most. severely attacked, metallographic specimens

17 from both loops, are compared in:Fig. 2.1.11,
Further tests are being made to determine the
cause for greater attack in loop 30 than in loop 39.
The metallographic examinations did ‘not reveal
any positive signs of carburization in either loop.

Chemical cnalyses of the fuel before and after
the tests did not show any significant changes as

a result of contact with the Inconel and the

graphite. For example, the carbon content of the
as-received fuel was 0.005%, and the carbon con-
tent was 0,018 and 0.024%, respectively, in the

fuel from loop 30 and 39 at the concluswns of the
- tests, . '

~ Chemical analyses for carbon were also made on

~each of three successive OOIO-m.-deep ‘millings

from the inner walls of specimens taken: from the

same locations as the metallographlc :specrmens.-

56

" CUNCLASSIFIED -
Y-18854 -

" UNCLASSIFIED
ORNL-LR-DWG 27123

 

 

 

" HOT LEG

 

 

e

-i—-—FLOW
COLD. LEG J‘

-

 

 

  

3 NO DATA

 

]
¥
9
/
1

" SCALE © 4 2 3 4 56 7 8 9 10 INCHES -

Fig. 2.1.10.. ."S'Ié'e“t»ch'of'j‘ilvicqne'l Thermal-Convection

Loop . with Gruphlté Insert in Hot Leg. Locations ot
which . speclmens were. rernoved are indicated by

circled numbers. - Numbers opposufe the curcled numbers

‘outsule fhe loop are attack, in ‘mils, in loop 30;
_ numbers ms:de the ioop are ofluck in |oop 39 '

The results did not indicate any si'gnifi‘can't ¢d:b6n'
_increase- in comparison with the carbon content of

as-received Inconel sampled in the same manner.

Dimensionally and Qisua“yr the graphite inserts

were unchanged by their exposure, except that the

CCN .graphite had changed from a glossy black to

a dull gray metallic color, The color change was :
uniform: on the inner surface but mottled on the

exterior. Photomicrographs of specimens of the
graphite inserts before and after the tests are
shown in Figs. 2,1.12 and 2.1.13, Fuel was found
in the pore spaces throughout the ‘/4-in.-fllick walls

of both inserts after the tests; however, the C-18

graphite did not appear to be attacked. There was

essentially no attack on the: CCN graphite, but it
"had a thin metallic-appearing film on the inner
and outer surfaces. The film was too thin and
erratic for accurate thickness measurements to be

 
PERIOD ENDING JANUARY 31, 1958

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

     

 

§ UNCLASSIFIED
- ¥24888
Fig. 2.1.11. (a) Specimen from Position 17 of Loop 30; (b) Specimen from fi’osition 17 of Loop 39. Etchant:
) aqua regia. 500X, ' ‘ ' ' ’ '
84 UNCLASSIFIED
z
0,02
o’.és
;',;
o
3 .8
Fig. 2.1.12. ‘C-18 Grupl-ufe é) As‘Rec;ivrecf finlifll"fi(‘b)r.AfffAer.,'E'xrfiézsfiré"-fqr 695hr to Fuel 30 af 1300°F il;n"dn Inconel
‘Thermal-Convection Loop. In (a) the gray material in some of the pores is plastic mounting material. In (b) the
\ | dark-gray material in some of the pores is fueli 100X, o ‘ '

57

 

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

1) 0.02

mc_HEs-':-

 

4lo.03 "

 

 

 

Fig. 2.1.13.
Inconel Thermal-Convection Loop. In () there is a thin metallic film at the surface. 100X.

made; it appeored to range from zero to much less

than 0,0005 in, in fhlckness. In scattered regions
the film penetrated into the specimen along the
graphite particles; the maximum penetration of
this sort was to a depth of 0.005 in.

A microspark analysis of the inner surface of the

'CCN graphite insert showed that the main com-
ponents, in the order of decreasing quontity, were
zirconium, chromium, and uranium and that there
were traces of sodium ond nickel. It is to be
noted, in particular, that chromium and nickel were

detected along with the cations of the salt. A

quantitative -analysis of millings from the graphite
'surface will be made in order to determine which
of these five components are from the metallic-
appearing film and to determine the quantities
present, The results of these thermal-convection
loop tests indicate, however, that at 1300°F the
corburization of Inconel in such systems' is
probably a slow process, if it occurs at all,

Static capsule tests at 1500°F have also been
made in the study of the effect of carbon on Inconel
and INOR-8 (heat SP-16) in sodium, in NaF-BeF -

"UF, (53-46-1 mole %, fuel 123), and in LiF-BeF, -

UF, (62-37-1 mole %, fuel 130). These 100-hr

58

 

 

‘CCN Graphite (a) As Received and (b) After Exposure for 1000 hr to Fuel 30 at 1300°F in an

static tests at 1500°F were primarily to determine
which material, INOR-8 or Inconel, carburized most
readily when exposed to graphite under the same
conditions and to .compare the carburization rates
of the materials in fuels and in sodium.

The specimens used in these tests were small
coupons, 40 mils thick, that were held in graphite
sleeves in Inconel capsules. The metal speci-
mens and the sodium or fuel were loaded into these
capsules under a purified argon. atmosphere and
then sealed by inert-arc welding. ‘ :

Metallographic examinations indicated, as shown

in Table 2.1.3, that INOR- 8 is more susceptible to

carburization than is Inconel under- these condi~
tions. - This is due fo the molybdenum in the

INOR-8 formmg carbides more easily than fhe'

chromium in Inconel, For example, it i$ known

that type 316 stainless steel, which contains 2 to

3% molybdenum, carburizes more readily than
similar austenitic stainless steels which do not
contain. molybdenum.?2  The appearance of the

INOR-8 affér the test is shown in Fig." 2.1.14.

 

2C. A. Zapffe, Staznless Steels; An Elementary Text
for Consumers, p 268, The American Soc:ety for Metals,
Cleveland 1949.

 
¥

*¥

i

PERIOD ENDING JANUARY 31, 1958

Table 2.1.3. Resulfs of Corburizut:on Tests of Inconel and INOR-8 Speclmens Exposed to Sodlum and to Fuel
at 1500°F for 100 hr

. Totol ‘carbon in as-received Inconel 6.038%

" Total carbon in as-received INOR-8: 0.020%

 

Alloy Weight - Total Carbon

 

Test Syetem Tested Change Content Metallographic Results
‘ - @ (@
Graphite-Sodium fnconel +0.50 0.60 Carburized to a depth of 10 to 12 mils, the first 1.5 mils
' ' “being a light band in which carbide particles were
agglomerated
INOR-8  +0.36 0.36 Carburized fhroughout;.firs'r 9 mils showed a heavy pre- -
cipitate; carbide particles less dense in interior of
specimen : |
Graphite—Fuel 123  Inconel  +0.20 0.04  Attack to maximum depth of 2 mils in form of small sub-
oo : o o L ' 'su'rche voids; no carburization observed
"INOR-8 +0.30 0.01 Small str.ingers 0.5.mil deep along edge; fine, scattered
' ' parricle formation to a depth of 1 mil
Graphite ~=Fuel 130 _!ncofiel ~0,37 0,03 Attacked to o moximum depffi of 2 mils in the form of
. _ small subsurface voids; no carburization observed
INOR-8  +0.03 0.01 Particles found in the grain boundaries to a'depi.'h of

1.5 mils

 

Chemical ahalySes did not show carburization of
either Inconel or INOR-8 in the -fuel mixtures.
Corrosion attack by fuel 130 was more severe,

in all cases, than attack by fuei 123, Similar

static tests were also conducted at 1250°F,

' Metallographlc examination and ‘chemical’ anulysis;{
" of these samples are not’ yet completed,” INOR-8-.
tensnle test specimens are. belng carl:urlzed by -
"usmg the sodlum-graphlte system, und fl'le specs-"':'
mens will be tested in order to determine the effect ;_:_‘_fj’

in both the fuels for 2000-hr periods at 1200°F. Both
_alloys were also tested in fuel 130 for 500 hr in
o seesaw-furnace apparatus with a hot-zone tem-
~perature of 1200°F, Hastelloy W capsules were
used as container materials in these tests be-
_cause “INOR- 8 was not available. The capsules

,.’_jr:were louded under a punfied argon atmos;:here with
-~ the ‘specimens ‘and enough fuel to fill the capsules -

- 1o one-third their volumes; the capsules were 1hen -
seoled by Heliarc we!dmg. R :

o of carbunzchon on tens:le strength ond elongflf“’“-;f

Chenmical ana!ysxs of the. fuel 107° to whlch the
_. . 82% Au-18% Ni alloy was. exposed showed 57 ppm ;

C fB All B F IS h . gold,” ‘and ‘the | cnalysas of the fuel 107 to which .
orroswno razmg oys y reel > : - the 80% Au=20% Cu alloy was exposed showed
© 123 ppm “gold. - No attack was observed on either

: D H Jonsen i
The precnous-metcl base bruzmg al!oys, 82% Au—- “alley in’ any . of the fests. The 82% Au-18% Ni g

 

+y

 

 

18% Ni and 80% Au-—20% Cu, which are being con
- sidered for use in the fabrication of fuel-sait—to-fr_
o coolant—suh heat exchongers, have been corrosion

tested in NaF-KF LiF-UF, (11. 2-41-45.3-2,5 mole -

- observed on the same alloy when tested in fuel

%, fuel 107) and in LIF-BeF ,-UF, (62-37-1 mole %,
fuel 130). - Static tests were run on both the alloys

- alloy - specimen " that was tested ‘in fuel 130 for -
500 hr in the seesow-furnace apparctus is shown in-
- “Fig.- 21,15, The thin layer on the surface was

107. X-ray analysis on this layer indicates that
it is composed of approximately 70% Ni—30% Au.

59

 
 

 

 

MOLTE'N-SALT REACTOR PROGRAM PROGRESS REPORT

Fig. 2.1.14,
100 hr ot 1500°F. Etchant: copper regia. 150X,

Cortosnon of INOR-B Welds by NaK and by
~ Fuel Salts

D. H. Jansen

INOR-8 plates from heat SP-16 that were welded
with various nickel-molybdenum-base welding rods
have been corrosion tested in seesaw-furnace
apparatus at a hot-zone temperature of 1200°F in

60

 

UNCLASSIFIED
Y¥.24915

Lol
po 0

. Y,
LhE

-
e "1,

A
e

-3"‘_‘ .

INOR-8 from Heat SP-16 (a) As Received and (b) Aher Exposure to o Sodwm-Graphfle Sysfem for

NaK (56-44 wt %) and in fuel 130 for 500 hr,
These tests were conducted- because the weld
metals contained some minor constituents that
have shown limited .corrosion resistance to these
mixtures; however, no attack was found on any of
the welds. The compositions of the welding rods
are given in Table 2.1.4, along with the observed
weight changes and the results of metallographic
examinations of the tested welds, :

v

h

 
L3

ay

 

PERIOD ENDING JANUARY 31, 1958

      
  

UNCLASSIFIED
Y-25064 '

 

 

 

 

. - LY Y ,- S0 » .
fi'* ‘ \ a ‘. ‘ -. xR " - v " 4
b"»g‘ ‘.; *‘. 1 ' "‘ L% ,'-
o e ""f‘*“‘“" R, sy el Sl oo
r':fi?azfi’u"'{f}'_'z"" ¥ u 0¥ 73w’ Se ST re

 

 

 

 

 

 

   

 

 

 

3 .
.' oot & .
TN IO oo AR AR E Lo
Beb i SN A g S e 1
etk o gpot Ay L
Sy .-‘fiff,v'. BN s T L O
- "Ql‘\"' ” '20.."‘ s ’.
AT R TE IR T L
"§'~‘:0f$¥,’3 J ¥ ,‘l_ + B "‘;"" .t, o g ." 2 ; .:..‘f -
[ SRR QO R, (0 ot vyt St L BT ]
"‘.’:.‘ j ‘. "."")-;., .,‘-'g‘- a%." & !3 ‘a ,..'.': - ... O‘. v “‘ OIZ
e d Ay £ o g . Byt ¥, e e
b a8 -"&"?4‘- S8 T .*.;,-‘f’f-:el‘ o0 6o & o
. ! . - -, - gt & . o
5 g At AVPR Y ..,“-‘3-,'3,.».;.-3-(4}'\ e Ny
Tjd\ s’ ’ ‘mg.g'fi o."'\ ',"2' ‘:::-95“ u .’ "':é Ci4
. L . - - y r - & Al :..:. <%0 f;"..' o~ o~
o i a3 :,fl >:'s!\~: Y Nt et i %
AN e g T s TR NN LG, T R T s e
Yoo \ ¢ ¥yt -1 Coiaatis ey y- (il e Y P e Ry L ARY -
Paig ot I 0o s o0& denr w"!{ 3 & ML P SRR Ly L
e 8 G i X .f',»'l- A% L 1, S A RS ke doil o
"?';."'..;{"4 Jw;f dp: ’, m:: ot ‘..»‘ %{,‘\;___‘;...: v 3/. . :“:..3..
T e A T AR RN T AR
2 Ty Lxts n... A e a st N fy. AWM Y -

Fig. 2.1.15. Speclfien of en 82% Au-18% Ni Brazing Alloy After Exposure to Fuel 130 for 500 hr in Seescw-.
Furnace Apparatus with a Hot-Zone Temperature of 1200°F. Etchant: KCN. 250X.

Table 2.1.4. Results of Corrosion Tests of INOR-8 Welds in NaK and in Fuel 130

Time period: 500 hr
-Hot-zone temperature: "1200°F

 

 

 

g Specimen 'T-estéd"-ifi Nol(.-_- RO Specilfiéh Tesfed in Fuel 130
: Nommul Composi!uon of : e e L - - e .
A F:Her Rod ~ Weight Change = - Metallographic _, Weight Change = Metallographic
TR - (®) - _oNotes " (W . Notes
74% Ni=15% Mo—6% Cr-S% Fe = +0.015  Noattack; cracks Iin . -0.035 No attack
(heaf 30-38 INOR 8) . - = . heat-affected zone S
70% Ni-16% Mo—?% Cr—S% Fe 0,035  Noarockicracks'in  —0.037 N attack; cracks
~“(heat SP- 16 lNDR-B) 0. < heat-affected zone - .. - - in welds .
”“60%N..25% Mo—7% Fe-—S% G- 40016 No attack on weld 0047 No attack on weld

2.5% Co (Ha sfellpy W)

 

61

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

PHYSICAL PROPERTIES OF INOR- 8-
T. K. Roche - '

Measurements were made of the modulus of

H. Inouye

elasticity, the thermal conductivity, and the tensile

properties of several commercial air-melted heats
of INOR-8 fabricated at Haynes Stellite Company
or at the Westinghouse Electric Corp. The compo-

sitions of the heats studied are given in Table .
2.1,5. The SP designations indicate material from-

Haynes Stellite Company and 8M designates
material from Weshnghouse.

Preliminary | thermal conductivity data for heat
SP-16 in the annealed condition are presented in
Table 2.1.6. The data wére obtained in vacuum,
~and Armco iron was used as the standard, Values
are glven for Inconel for comparison.

'The values of Young's modulus glven in Table
2,1.7- were obtained by senic methods.®  An
annealed bar of heat SP-19 was used.

Studies were initiated for determining whether
INOR-8 has a tendency to embrittle in the tempera-

 

3Datu obtained by S. Fulkerson of ORNL wnth sonic
equipment ot the Bureau of Standards.

Table 2.1.5. Chemical Analyses of INOR-8 Heats

ture range of 1000 to 1400°F. Specnmens are being

' 'uged for perlods of 0, 500, 1,000, 2,000, -5,000,
“and 10,000 hr.
- properties obtained. after 500-hr aging heat treat-

‘The " room-temperature " tensile

ments at several temperatures are presented in

" Table 2.1.8. These data show that the alloy does
not become embrittled during a 500-hr aging period,
- The higher strength and lower ductility of heat

8M-1 in comparison with heat SP-19 are ascribed

to a difference of 0.08% in the carbon content.

" Table 2,1.6, Thermal Conductivities of INOR-8 -
"{Heat SP-16) and Incqnel at Various Temperctures

 

Thermal Conductivity

 

 

Tem;;:(r:fl;fl-!re ,[cal/Cfn2°sec-(°C/cm)] o
INOR-8 Inconel
100 0,023 - 0.039
200 0028 0.041
300 0.042 o 0.043
400 0.050 | 0.045
500 0.059 - 0.048
600 0.067 0.050
(extrapolated)
700 0.075 0,052

- {extrapolated)

 

 

~ Amount Found (wt %)

 

Table 2,1.7. Modqlus of Elasticity for INOR-8 at
Temperatuies Up to 1050°C

 

 

Element - _
Heat SP-19 Heat 8M-1 "_Heaf SP-16
Mo . 16.65 16.20 15.82
Cr 7.43 7.47 6.99
Fe 4.83 6.1 4.85
c 0,06 0.14 0.02
Si 0.04 0.21 0.32
W Trace * 0.35
Mn - 0.48 0.69 0.34
P 0010 0.009 0.009
‘s 0,015 0.006 0.014
Cu o 0.02 * . 0.03
vV .10 . * *
B S * 0.04
Co - 0,51 o+ . 0.51
Ti . o o

Ni 7000 - 70.6 . 70.50

 

 

\ *No'_f dna[yzed f_'or.

Temperature Young's Modulus
(°C) (psi)
Cx 108
14 L 3.7
223 o 29.3 -
412 o 27.8
501 27.1
576 26,3
636 262
701 . 24.8
800 = - 23.7
857 ' 22.7
902 219
953 : 20.7
1000 , 194
1050 | 7.7

 

AL ]

 
 

 

i

ap

L1

 

PERIOD ENDING JANUARY 31, 1958

Table 2,1.8. Room-=Temperature Tensile Properties of INOR-8

 

Yield Strength at

 

 

Tensi:e S.;rength 0.2% Offset EIor(n;c;tion
si %
Heat Treatment P (psi)
Heat Heat Heat Heat Heat Heat
SP-19 8M-1 SP.19 8M-1 SP-19 8M-1
Annealed 114,400 117,100 44,700 51,900 50 39
Annealed ond aged 500 hr at 1000°F 112,000 115,700 | 42,500 47,200 53 43
Annealed and aged 500 hr at 1100°F 112,600 114,500 44,000 48,000 51 43
Annealed and aged 500 hr at 1200°F 112,300 114,600 44,700 48,600 51 43
; Ar'meo’ed and uged 500 hr at 1300°F 112,000 113,400 44,500 47,600 49 4
Annealed and aged 500 hr at 1400°F 112,400 116,000 43,900 47,000 50 40

 

MECHANICAL PROPERTIES OF INOR-8
- D. A, Douglas
Tests are under way for obtaining the basic data
on the strength of INOR-8 required for design

calculations. The confidence level which can be
applied to the data and the varicbles which affect

~ the reproducibility of the data will be defined,

Studies will also be made of the behavior of the
metal under both static and dynamic loadings in
order to more accurafely predict fhe service life
of various component parts,

Since relatively long pericds are required to
obtain. data on ploshc properties, preliminary data

on fensile properties are being. obtained for use -
~in design studies.  The yield srrength at 0.2% - 45 be mVeshgated " Creep tests are therefore

“offset and . the rupture strengths -of _INOR-8 were - under way at stresses. ‘of 12,000 to 30,000 psi with

 measured - in the temperature range of 1000 to
~ 1300°F ond at room temperature, The results of
'-_the meosurements are summorlzed in Table 2. 1 9 L

o The dora presented in Table 2.1.9 were obtamed- -

~-on sheeir specimens .and must be consudered as -
- '_opprox:rnote because of expenmentai errors in the
~ elastic-portion of the stress-strain data,  Errors’
‘occur _partly ‘because it -is difficult to. ochreve--i
accirate alignment with a sheet. specimen and
'pcrfly becuuse subsize’ $pecimens are sensitive to .
_ expemr-eniel vonohons._ Conventional specimens

0.505 in. in dlumerer are being machined from a
wrought bar so that more accurate values can be
obtained.

The study of the plastic properties of INOR-8
is being made in order to determine whether INOR-8
will deform plastically under reactor operating
conditions. In the relaxation tests used for this
study, a specimen is loaded to a fixed amount of
strain and the resulting elongation is maintained
either by adding to the load or by subtracting from
it. The need to remove the load to maintain the
fixed strain would indicate that the material de-
formed plastically. The results of a series of
tests at 1200 and 1300°F are summarized in Tabie
2.1.10.

The large decrease in stress with time indicated
that the plastic properties of the INOR-8 were
important and that rhe creep strength would have

the specimens exposed to fuet 107 -at 1100, 1200,

~and 1300°F, The status of these tests is pre-

sented in Toble 2.1.11,
Tests were also performed at femperotures con-

‘gislderob!y above the anticipated operating tempera-

ture of the reactor as a means of gaging the damage -

to the metal which might occur through occrdentol

rremperoture ‘excursions. - These - resulfs, are .pre-
-sented in Fig. 2.1.16. '

“Ancther series of resrs will be run in order to

;obrom a siorrshcol estimate of the reproducibility

of‘ creep results from one test to another and from
one heat of material to another. These tests will
be conducted in air at 1250°F,

63

 
 

 

 

 

 

MOLTEN-SALT REAC_TOR PROGRAM PROGRESS REPORT

Table 2.1.9. Elastic Pr_operties of INOR-8

 

Yield Strength at

Ultimate Strength

 

' ‘ Temperature Ductilit
INOR-8 Heaf | :’:F) 0.2?p:)i|;fs_et o) - e Y
Haynes SP-16 Room 45,000 106,000 58
Haynes SP-19 Room 45,000 ) 114,000 39
 Westinghouse " Room 52,000 117,000 50
‘Haynes SP-19 1000 27,000 90,000 e
 Westinghouse 1000 36,000 100,000 o
Haynes SP-19 1100 29,000 93,000 50
Westinghouse 1100 38,000 103,000 ) 37
" "Haynes SP-16 1200 25,000 67,000 44
" Haynes SP-19 1200 27,000 82,000 36
Westinghouse 1200 38,000 83,000 16
Haynes SP-16 1300 24,000 58,000 v

- Haynes SP-19 1300 28,000 70,000 24

Westinghouse 1300 | 38,000 70,000 ST

 

Table 2.1.10. Relaxation Data for INOR-8

 

 

 

Tempérotufe Strain Initial Stress Stress (psi) for Constant Elongation
(°F) . (psi) (psi) At 1 hr " At10hr At 100 hr
1300 © 0.05 11,000 11,500 10,000 6,000
0.1 21,500 21,500 16,000 5,500
0.2 . .29,750 . 20,500 - 10,500 , 4,500
1200 0.05 12,000 12,500 12,000 10,000
0.1 22,500 23,000 22,000 - 17,000

 

WELDING AND BRAZING STUDIES
' - G, M. Slaughter

Metal Seals for Remote-Disconnect Flanged Joints

In the development of mechanical joints that can
be disconnected by remotely controlled mecha-
nisms, it has been necessary to investigate various

‘types of seals. Tests have been made of the

feasibility of heating a sealing material in an

‘annulus to make or break the seal or of pouring

molten metal into a preheated and cldmped flange
assembly, The results of mockup tests of such a

64

joint, designated “‘cast-metal-sealed flange joint,"”
I gn ge | _

are presented in Chap, 1.2, 7

A survey of phase diagrams of possible seal and
flange materials which should be relatively
immiscible in each other indicated that the silver-

nickel system might be useful. Silver and iron
~were found to be virtually insoluble in each other,

and copper and iron possess only limited solu-
bility in each other. It also seems probable that
the silver-copper eutectic alloy, Handy & Harman

alloy BT, which melts at 1435°F, could be used
with iron-base flange materials,

v

AL ]

 
 

wh

o

Toble 2.1,11. Creep Data for INOR-8

 

Temperature Stress  Strain Time Creep Rate

 

(°F) {psi) (%) (hr). {%/hr)
1300 30,000 15.33 110 Ruptured
25,000 6.53 187 *

20,000 9.56 882 Ruptured
15,000 10.99 2894 Ruptured

1200 30,000 4.7 9**  Ruptured
25,000 2.84 1195 2x10-3
20,000 2.70 1894 2 x 10~4
15000 0.81 1863 2% 104
12,000 0.97 1172 6x 103

1100 25,000 0.70 164 *
15,000 - 0.58 68 *
12,000 0.3% 43 -

 

*Test under way.

 **Tegt to be repeated; low elongation at failure is not
considered to be typical.

. UNCLASSIFIED
ORNL—LR—=DWG 27909

[ ' A
"-1650‘F / )

—{500°F

 

2

 

S

 

 

STRAIN (%)
o
>
o
Q
L
\%,\

606095\ /
P

 

N,
N
R

 

 

 

 

 

 

 

 

 

 

 

 

 

0 50 100 150 200 250 300 . 3/0 400 450 - 500
L ,TIME(hr) e

Flg | 2.1, 16.‘ Creep Curves for Soluiion-Annealed. )
‘ INOR-B Tested in’ Fuel 107 ut Varlous Temperutures' R
: ;rond Sfl-esses. R o : , PR ‘

The speczmens ‘are shown in- Flg. 2.1.17.
Meta“ographlc ‘examinations were mode of dupll-

cate samples after solidification and after holdmg -

at elevated temperatures for extended periods of

time in order to determine the extent and type of -

PERIOD ENDING JANUARY 31, 1958

diffusion. The excellent wetting of nickel by
silver in dry helium is indicated by the small
contact angle shown in Fig. 2.1.18, An even
smaller contact angle was found on the nickel|-BT
alloy sample. Metallographic examination of the
silver-nickel interface revealed no penetration of
silver into the nickel after brazing and after sub-
sequent aging for 500 hr at 1200°F. Only slight
penetration of nickel by the silver-copper alloy
was observed ofter aging. The marginal wetting
of INOR-8 and Inconel by silver in dry helium is
illustrated by the contact angle of approximately
90 deg shown in Fig. 2.1.19. The BT alloy ex-
hibited very poor wetting on both INOR-8 and
Inconel. The wetting of iron by silver and the BT
alloy in dry helium was poor and intermittent,

The deposition of an electrolytic nickel plate on
Inconel and INOR-.8 samples was found, however, .
to promote good wetting by both silver and the BT
alloy. The nickel plate was not penetrated by
silver, but there was some solution of the nickel
plate by the BT alloy. The deposition of an
electrolytic silver plate on the electrolytic nickel
plate did not improve the wetting by either alloy,

Aging experiments at 1200°F are now under way

on electroplated materials.

A preliminary sealing test specimen was de-
veloped that consisted of a nickel tube and a
nickel container in which the seal could be made
with silver under a dry helium atmosphere. The
joint was made and broken four times, and the
joint was leaktight after each sealing. Thus there
appears to be excellent wetting of nickel by silver
under repeated heating ond cooling cycles. As

~ stated above, mockups of cast-metal-sealed flange
~_joints ‘are bemg tested fhat uhhze the results of
 these studles. -

Weld ing of INOR-8 Tubing
G M. Slaughter
Before the fabncahon of INOR-8 tublng from heat

7 A series of tests was fhen conducted in d,-yi - SP-16 into the various corrosion loops of interest
- hehum in order to study the wetting character;st:cs‘,
of silver and the silver-copper eutectic ailoy on
“nickel, iron, and other possible flange materia)s — - -

Vr'type 316 stamless steel, INOR-8, -and Inconel ‘

was mmated, it was decided that a metallurglcal

iinvestigation of typical welds should be conducted
A preliminary mvesngahon, described prevnously, _
?_‘_mdtcated that fhe use of SP-]6 filler wire for the
_welding of /2 in. plate from heat SP-16 was, in
"general, | ‘unsatisfactory because of weld-metal

cracking. Also, welds made on similar specimens

 

4P; Patriarca and G. M. Slaughter, MSR Quar. Prcg.
Rep. Oct. 31, 1957, ORNL-2431, p 18.

65

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

SILVER '
BRAZING TEMP. 1800°F INOR 8

HELIUM ATMOSPHERE

BT SILVER SOLDER i
- BRAZING TEMP. 1480°F

HELIUM ATMOSPHERE 70N

  

SP16

~ .UNCLASSIFIED
T Ya24465

    

stess - INCONEL . NICKEL

I T

 

36 SS - INCONEL -~ NICKEL

 

Fig. 2.1.18. Specimen Showing Excélle:rlrWeflihg of a
Nickel Tube by Silver. Unetched. 50X. Reduced 30%.

under . condmons of high restrumt resulfed in
severe bcse-metal cracking.

Test welds have since been made on fhe / in,-

OD, 0.035-in.-wall tubing from heat .SP-16 under
low-restraint conditions.  Since the previous
* experiments had indicated that SP-16 filler wire
was not satisfactory for this application, material
from an ORNL heat bf INOR-8 (heat 30-38) which

had shown promise in weld-cracking tests was

fabrscated mio wire and utilized as flller metal in

66

. :Fig.' 2.1.19. Specimén S}‘:c.'w'ing.'-flarginq-l Wefling of
INOR-8 by Silver. Unetched. 50X. Reduced 30%.

~ the brodbc_tion of the test welds. Heat 30-38 has

the nominal composition 15% Mo—6% Cr—5% Fe-—
0.5% Mn=0.5% Al-0,06% C—bal Ni. - A photomicro-
graph of a typ:cal welded |omt is shown in Flg.

2.1.20.

“The results of visual, rudlographlc, and metallo-r
graphic examinations .and mechanical .tests at
room temperature and at 1300°F on as-welded-

.specimens indicate -that sound, low-restraint butt

welds of SP-16 tubing con be made with heat 30-38

 
o

e

 

e R ALl b

 

 

PERIOD ENDING JANUARY 31, 1958

UNGLASSIFIED
«24329

Fig. 2.1.20. Typical Joint in SP-16 Tubing Welded with Heat 30-38 Filler Wire. Etchant: chromic and hydro-

chloric acids. 25X,

filler metal. No base-metal or weld-metal cracks
were found, and the room- and elevated-temperature
mechanical tests of the as-welded joints indicated
satisfactory characteristics. The properties of
the joints after aging ot elevated temperatures are
being investigated as a means of determining their
over-all suitability for high- temperafure applica-
tions,

Since no significant qucmhty of heat 30-38 flller

, Numerous saddle’ welds were exommed durmg,p
the development of the weldmg procedures, and "
infergranular - base-metal ‘cracks fo the extent - of
20% of the tube wall thickness were oceasionally i

observed in the microsections,

The higher re-

straint conditions involved in the fabrication of

meicl ~was -ovailable in the optimum. wire size,
' maierlol from ‘another ORNL heat (30-72) of the -

some “nominal - composutlon was - processed into-
~wire, Sample butt welds of the /B-m.-OD 0.035--
“in.awall $P-16 tubing were ‘made as “before, and a - -

. ‘were described prevnously.‘l_
preliminary evaluation made to ‘obtain data for the

‘ - Ctest, whlch -utilizes ‘an inert-arc fusion. pass on a
~“development of a procedure specification indicated” " -3 '
~ -that the welds were sut:sfactory. _
“mental work has been. completed on which to base
" a procedure specification ‘and an opemror s quoll-j"

L flcahon test spec:hcuhon. . T

‘The develop-;f_-‘

these welds undoubtedly explain the presence of
base-metal cracks in the saddle welds aond the
freedom from cracks in the lower restraint butt
welds. Since the detection of these defects cannot

be ‘ensured by radiography or dye-penetrant in-

spection, it is recommended that the SP-16 tubing

'}nof be used for critical applications, such as in-
pile loops,

Evnluutlon Tests for Welds -

Prehminary results in the development of screen-

,-'_._mg tests for determmmg the relative susceptl-
' ‘-l)llll’les of various alloys to weld-metal cracking

‘The circu lor-groove

’,,szqn.-wude, 2-|n.-dta, }g-ln.-deep clrcular ‘groove
~machined 'in-a "4 % 4-in, specimen of /2 in, plate,
“has now been used fo test the weld-metal cracking
"tendencues ‘of the - INOR-8 alloys Westmghouse‘
8M-1-and Haynes SP-l9 {see Table 2.1.5, above,

for- H1e composmon of these alloys).

"No weld-

__i'metol “cracks -were. found in either: alloy, the
"_'-Weshnghouse 8M-'l spec:men is shown in- Fig.
2121, |

Metollographic 'secfi_oning_ of this type of speci-
men can also be used to determine the susceptibility

67

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

of the material to base-metal cracking during weld-
ing. Base-metal cracks |dent|ca| to those found
in htghly restrained test welds? ‘were found in a
Haynes SP-16 specimen,-as shown in Fig. 2,1.22,

 

Fig. 2.1.21, Weshnghouse 8M-1 Circulur—Groove Weld-
Crockmg-Test Specimen.

 

 

‘No base-metal defects - were observed in the

Westtnghouse 8M-1 matenal “a typical area’ of

which is shown in Fig. 2.1 23. ‘No strmgers or
“inclusions are evident. Ahhough no ‘base-metal
~.cracks .were found in the Hoynes ‘SP-I9_specnmen,

the presence ‘of slight, occasional, fusion-line
porosnty was noted Thls condmon |s shown in

‘Fig. 2.1.24,

A crack test for materials avullable onIy in rod
form was also deveioped “A longitudinal slot,
3/16 in. deep and / in. wide, is machined in ¢
z-ln.--dla bar and a fusion weld is made along the
slot. Weld-metal cracks were found in a Haynes
SP-16 bar, but there were none in c Westmghouse
8M-9 bar. The two specnmens ‘are. shown in

‘Fig. 2.1.25.

The weldmg charocienshcs of seven’ dlfferent

- heats of INOR-8 from three different sources have

been studied, and ‘only the Huynes SP-16 heat

“exhibited crccklng tendencies. This crockmg has
' been ‘attributed to the melting practice and should

not be considered as mdlcahve of fhe properhes

of the alloy.,
Weld test plates are bemg prepared as a means .

of evaluating the weldability of heavy sections of
nickel-molybdenum alloys. Three stages in the

 

 

 

W.mfim;rnwm
UNCLASSIFIED,
- Y«24908
0,
p- b J
x
o -
z
0.02
0,03
x
O

 

Fug. 2.1.22. Buse-Meta! Cracks Found by Metuilogrophic Examinoflon of a Huynes SP 16 Curculcr-Groove Weld-
Crackmg-Test Specimen. Etchant: HCl + CuCl + alcohol. 100X.

68

 
 

 

 

 

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED|
Y24910

S

 

INCHE

 

0.02

 

 

 

Fig. 2.1.23. Typical Area of Westinghouse 8M-1 Circular-Groove Weld-Cr‘acking-Tes.t‘Specimen. Etchant:
HCI' + CuCl + alcohol. 100X, - . A

UNGLASSIFIED |

 

INCHES

iy

6.02

 

 

 

 

e

 

 

ng. 2.1.24. Haynes SP-19 Circular-Groove Weld-Crccldng-Tes! Specimen Showing Slight, Occasional, Fusion-
Line Porosity. Etchant: HC! + CuCl + aleohol. 100X.

69

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

 

~ UNCLASSIFIED
O Ye24776

Fig. 2.1.25. Rod Weld-Cracking-Test Specimens Showing a Crack in a Haynes SP-16 Bar ond No Cracks in o

Vestinghouse 8M-9 Bar.

preparation of a typical test plate are illustrated
in Figs. 2.1.26, 2.1,27, and 2.1.28. These plates
provide specimens for mechanical property studies
of welded joints ond are useful for radiographic,
metallographic, and hardness studies, as well as
for obtaining general information pertaining to the
welding characteristics of the materials under
conditions of high restraint, A summary of the
status of the evaluations of the INOR-8 weld test
plates made to date is presented in Table 2,1.12
" For hardness studies, samples of the welded
test plates are removed and measurements are
made on the weld metal ond base metal in the
gs-_&vélded “condition and after aging. ' Preliminary
work has consisted of determining the Rockwell B
“hardniess of the two zones, but complete hardness
traverses across the weld heat-affected zones are
"being made to obtain a more complete understanding

70

of the influence of welding. The results of the
preliminary hardness measurements on test plates
38 and 40 are presented in Table 2.1.13. in com-
parison,  the hardness of Hastelloy B and W weld

metals in the as-welded condition is opproxlmately '

20 on the Rockwell C scale, which is approxi-
mately equal to a value of 98 on the Rockwell B
scale.. After aging at 1300°F for 200 hr, the hard-

ness of Hastelloy W weld metal rises to approxi-

mately 32 on the Rockwell C st’:ale,lrwh_ileithat of

Hastelloy B rises to approximately 41,

It may be seen that heat 30-38 weld ‘metal is
slightly- harder in both the as-welded and aged
conditions ‘than Haynes SP-16 weld metal. How-
ever, no significant hardening that can be attributed
to aging is evident for either alloy, Both these
alloys are- consuiercbly softer after aging than
either Hastelloy BorW,

'y

 
 

PERIOD ENDING JANUARY 31, 1958

BT = T - \ : B UNCLASSIFIED)|

PHOTO 42156

   

       
   
  
 
  
 
 
  

 

 

] £
OA K HIDGE  HATIONAL L ABORATOR

lg:chTAgsgggy - Approxlmateiy 40 bend tests have been made on

' Hasfelloy W weld metal and heat 30-38 weld metal
at room ond elevated temperatures. These tests
‘were ‘made on.specimens - in the as-welded condi-
“tion ” qnd on : aged specmens._ The testing and
'-agmg temperatures used were room temperature and

1100, 1_200': ,1300 1500, and 1650°F, The data

 

 

 

 

      
    

 

   

; o '-shown in Fig. 2.1.29, The Hastelloy B mpple was

n - . R I ~ part of the finish-machined Hastelloy B pump
<Y Fig. 2.1.27. Weld Test Plate Assembly After Tack  barrel. The opposite end of the SP-16 adapter
- Welds Were Made. . - S was welded to the SP-16 tubing of the loop. Filler

  

 

    

 
 

 

 

MOLTEN.SALT REACTOR PROGRAM PROGRESS REPORT

. Table 2.1.12. Summary of Status of Evaluations af INOR-8 Weld Test Plates.

 

 

- Plate - Base Metal - | .. Filler Metal = L . . Status of Evaluation
: 36 . 'Hoynés SP-16 - Haynes SP-16 " Weld me_tfl cracked severely; no mech&hical
' ' S . (sheared strip) o _prppefly tests conducted v
37 ‘_ - Haynes SP-]6 Hasrielrloy' W | Plate 'm'c,l1c-:7hined into side~-bend specigfiens -

. for roégm- and elevated-temperature testing;

. tests 90% complete

38 Haynes SP-16 ORNL heat 30-38  Status same as for plate No. 37
39 Huyhe_s SP-|_6 (weld Hastelloy W : o 'Metallog';‘r'aphic evdluation'complefed; no
o ‘made under lower : K mechanical property tests to be conducted

restraint than

‘conventional test

plates)

- i'.AO -~ Haynes SP-16 _ : Ho.ynes 5P-16 rod : Weld n\e_tal cracked severely; no.tfiechanical
- e obtained from - property tests conducted - :
Haynes Stellite Co, ;

41 Haynes SP-19 Hayne; SP-'I9_ - ) qufe being mcchinéd into side-bend fest.

' (sheared strip) - . .. specimens v :
- 42 ~ Haynes S‘P-'|9 Hay‘nes SP-19 . Stafus same as for plate No. 41 |
- (sheared strip) - '
43 Haynes SP-16 Westinghouse Plate to be machined into side-bend speci-~
7 heat 8M.5* " mens
44 Haynes SP-16 ORNL heat 30-73** Plate to be machined into side-bend speci-
' mens T

 

*Nominal composition: 17.3% Mo-7.0% Crx5.2% Fe-0.087% C-0.19% $i-0.79% Mn-0.001% P-0.002% S~
' 0.018% Ti-bal Ni. ‘ o ' ’
**Npminol cpmposifion: 17% Mo—-8% Cr—-5% Fe-0.5% Mn-0.06% C-bal Ni.

Table 2,1,13. Hardness Data for Test Plates 38 and 40

 

‘Har.dness on Rockwell B Scale -

 

 

.+ Condition " - Base Material ~ Plate 38 Weld Metal "Plate 40 Weld Metal
S ' (Haynes SP-16) _ {Heat 30-38) (Haynes SP-16)
 Aswelded | 89 | 97 | 92
* Aged at 1100°F for 200 hr 89 a e 90
Aged at 13(50"’# for 200 hr 87 95 | 8
 Aged at 1500°F for 200 br 85 94 86
| Ag'edro'r i650°F for 200 hr 85 | 91 | B - 86 |

 

 
 

oot L

 

 

b

 

 

/

No evsdence of crackmg or fallure of the heat"

  

PERIOD ENDING JANUARY 31, 1958

 

 

Fig. 2.1,28. Weld Test Plate Assembly After Welding.

wire from heat 30-72 was used in the fabrication = occurred in a mockup experiment. In the mockup
of the entire loop. ' experiment, the components of the most highly
Metallographic examinations were made of sec- stressed portions of the weldment were assembled,

tions from the area of the failure. Profuse cracks

.were _found in . the Hastelloy B nipple, and the . . . - DT e
fracture interface was predominantly in fhe heat-r-f” e :
affected zone, as shown in Fig. 2.1 30 In some -
cases, cracks were found as much as- ./“5 in, away .
from the frocture xnterface, as shown in Fig. 2.1.31.

30-72 weld metul was_noted, . It thus appears that " -
" the failure can be attributed to the lack of ductility = =~
—of -the Hastelloy B mpple and its lnablllty to
" accommodate - the - sfresses encountered durmg,”—
stqrfup of fhls loop. :

 

Fcbncuhon of Test Components

 

Work hcs been :nmated on the fubrlcatton of two; S - : S : :
Haynes SF’-'|6 pumps for forced-c:rculat;on corro- - i Fig. 2. .29 Joint Thut Fnlled During |nl!'|u| Henflng
sion festing loops. A modified fobrication pro- »f First INOR-8 Forced-Circulation Loop, The break is
cedure is being utilized in the fabrlcqhon of these at the fusion line of the weld joining a Hastelloy B
pumps, since extensive base-metal cracking  nipple to an SP-16 adapter.

73

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

g
H
1
1

i
i
2

Fig. 2.1.30. Cracks in Heat-Affected Zone of Hastelloy B Nipple, Etchant: chromi
100X. |

 

S
e
v

F|g 2.1.31. Diverse Cracks in Hastelloy B Nipple. Efci-;anf: chromic and hydrochloric acids. 75:X.

<p e

¥

 

 

INCHES

0.02

 

0.03

 

 

 

c and hy_drochlori_'c acids.

INCHES

 

0.03

 

 

0.08

 

75X

 

 

0.08

 

 

 
43

 

ad

N~

“welded,

 

 

~and metcllogrophrcaliy exal‘mned “after
welding. The components studied are shown in
Figs. 2.1.32, 2.1.33, and 21,34 before and after
being welded. Metallographic sections -of the

finished welds revealed extensive _bcxse’-metcl_

cracklng.

In the modified fobncotlon procedure, a I-m. cap
is premachined to approximately the desired
dimension -in order to reduce the restraint on the

joint, and additional stress-relief anneals are

used. The two units being fabricated are uppI'OXI-
mately 25% complete.

The fabrication of a Haynes SP-]6 |NOR-8 in-
pile loop for operation in the LITR was completed,
as were nine INOR-8 thermal-convection loops.

UNCLASSIFIED
Vauls

 

Fig. 2.1.32, Pump Fubricution Test Cor_np_onents Prior

to Welding.

 

 

. UNCLASSIFIED
T Y-24450 -

Fig. 2.1,33. Top View of Completed Pump Fabrication
Test Assembly.

PERIOD ENDING JANUARY 31, 1958

 

T : " UNCLASSIFIED,
’ Yo24448

  

 

 

'Fi§§_2.l.34. Bottom View of Completed Pump Fa_bri-
cation Test Assembly. '

Forty-seven Haynes SP-16 hot ductility test
specimens were machined and sent to Rensselaer
Polytechnic Institute for evaluation. A report of
the findings will be submitted when the study is
complete. Haynes SP-19 specimens are bemg

mochmed for similar studues.

DEVELOPMENT OF NONDESTRUCTIVE
TESTING TECHNIQUES

R. B. Oliver
R. W. McClung
Evaluutron of INOR-8 Tubmg

Immersed ultrasound and eddy-current techmques, ,

J. W, Allen

jrt:tcllogrc:pl'uy, ond fluorescent penetrants were used, '
1o evaluate’ severol ots of INOR-8 tublng in sizes-
. that, included / in.. OD 0. 025-|n. wall;” /2-|n. 0D,
,'0.045-|n. wall; ‘and 0200-|n. OD 0. 050-|n. wall,
}fi,;Dlscrete discontinuities were" noted in: all the
- tubing.
_ being exummed metoilogrophlcolly. The lamination
-.* shown in Frg. 2.1.35 is typical of the more common -
o defects, it was defected by the immersed ultrasound

Mony of the . areas of discontinuity ‘are

techmque in '—h-m.-OD 0025-|n.-wu|| tublng.

 Similar * defects “were noted in the /2*In.'OD :
-0, 045-|n.-wc||| tubmg.
: f-t-*r_"/4 in. ‘and “z in. tubmg were. vo:ds olong the weld
bead, such as those illustrated in Fig, 2.1.36. No

Also common to ‘both the |

results - of metallographic examinations are yet
available for the 0.200-in,-OD tubing, but macro

75

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

 

 

 

 

Fig. 2.1.35. Lomination Found by Immersed Ultrasound Inspection of 1/‘
Etchant: chromic and hydrochloric acids. 100X.

H
[
ik
£
R

 

 

Fig. 2.1.36. Subsurface Void Found by Immersed Ultrasound Inspection of ‘/z-in.-OD, 0.045-in.-Wall INOR-8
Tubing. Etchant: chromic end hydrochloric acids. 100X.

76

 

 

 

I

 
 

 

 

 

   

JL“.

0

 

 

) exammahon of sechons of 1he ;fube“ revealed”
- several severe. lops on the inner. surface. Despite
’ ,_the formudcbie ‘appearance . of . the “illustrated de-

fects, -the over-ofl qualfly of the INOR- 8 tubing

" has been better than that of other commercial
mckel-molybdenum olloy tubmg prewously evalu- .
 ated. S '

Cluddmg Tl'uckness Measurements nnd -
Bond lnspechons _ -

Development studles ‘were conhnued of ?ech--;
_-niques “for the. meusurement ‘of claddlng thickness.
“with the eddy-current probe coil,
~ thickness will require a special adaptuhon of the
'techmques in order to- compensate for varlctlons
“in  configuration, ‘materials, and thlcknesses. '
- Basic mformohon on the parameters mvolved is
i bemg obtmned. : - o

Technlques are ulso bemg developed for ther
evaluqhon of bond quality by ultrasonic methods. -

As in the case of cladding thickness measure-
ments, “techniques “will be ‘required whrch _are

,pecullar to the parhcular appllcohon. S

Each claddlng

PERIOD ENDING JANUARY 31, 1958

lNSPECTIDN RESULTS _
G M Tolson ,J H. DeVan

Muienal Ins pechon

A summary " of material anspechons during the
quarter is presented in Table 2.1.14. The infended

‘usage was the criterion used to determine the type

of - inspection performed and the acceptability.

" Whenever possible the rejected ‘material was down-
- graded_ for less critical applications,

“The - over-all rejection rate for experimental

‘nickel-molybdenum tubing was high, but it varied
considerably from.lot to lot end, in facf -wdas
‘quite low for some lots. Most of the rejections
~ - were due to defects in the weld areq, such as lack
:'of fusion (Fig. 2.1.37), cracks (Flg. 2.1.38), and
_pores (Fig. 2.1.39). The seamless tubing was in

generul of better quality than the. Weldrawn tubing,

and _the Weldrawn tubing was~ of beh‘er quality
“than. fhe us-welded' ‘tubing. A casting made from.
“INOR-8 was re]ected because of large pores, but

the general appearance of the casting indicated
tho_t casting of complicated shapes may be possible
with improved techniques.

. Table 2.1.14. erteriote_ _ln'epeeted |

 

 

|fem 4 - ) U Mqterieli - - _Quunfi_ty inspec'fed- Qu.on'tify-Reiected .
Tubing  Experimental nickel-molybdenum alloys 3293 # 9 in. 2006 ft 8 in.

_ Tubing ... Inconel .

 

 

INO R-8 ca shng

lnconei Calrods

‘:tfi‘éiié. o

T 1h B

 

77

 

 
 

 

MOLTE.N-’S'ALT'VREKACTOR PROGRAM PROGRESS REPORT

   

E UNCLASSIFIED
N T-13719

 

 

 

Fig. 2.1.37. Specimen Showing Lack of Fusion Found
in Weld Area of INOR-8 Tubing. Etchant: aqua regia.
100X. Reduced 31%. ' : :

 

 

 

Fig. 2.1.38. Specimen.Showing a Crock Found in Weld
Area of Weldrawn INOR-8 Tubing. Etchant: aqua regia.
100X. Reduced 31%.

Weld Inspection

" Welds are inspected according to the quality
needed, and the three general quality levels con-
sidered at the present time are designated C, CN,
‘ond S, The C type of weld is of reactor quality and
receives visual, penetrant, and radiogtaphic inspec-
tion. The CN type of weld is just as critical as the

C type, but it is in a position such that it cannot

be radiographed. The S type of weld is used in
‘noncritical applications and receives only visual

inspection, The specifications are being revised

to. include a fourth type of weld which will be
designated NC, The NC welds will be used where
a leaktight joint is required but certain deviations

78

UNCLASSIFIED
T-14029

 

 

Fig. 2.1,39. Specimen Showing @ Pore Found in Weld
Area of As-Welded INOR-8 Tubing. Etchant: chromic
and hydrochloric acids. 50X. Reduced 33%.

can be allowed. The type of inspection of an' NC
weld will depend on its application.

The amount of penetration allowable on C and
CN welds made on tubing has been increased from
0.025 to 0.035 in., because there was difficulty
in maintaining such a tight tolerance and the test
loop work now in progress does not require such
rigid control.

A total of 131 C welds made on INOR-8 material
was inspected, and the rejection rate was 9%,
which is less than one-half the rejection rate
previously found for Inconel welds. A total of
248 C welds and 197 CN and S welds on other
materials was also inspected, with 57 C welds and
3 CN and S welds being rejected.

Failure Analyses

A Chromel-Alumel thermocouple which failed
after 240 thermal cycles between 800 and 1200°F
in NaK was inspected to determine the cause of
the failure., The thermocouple was encased in a
protective Inconel well, and therefore it was
necessary to locate  the failure by radiographic
inspection. ' A break was indicated on one of the
wires near the thermocouple junction, and metallo-
graphic sections were taken at the failed area.
The  thermocouple wire at the break had necked
down considerably, and thus a ductile tensile
failure was indicated. Since the junction of the
thermocouple had been firmly attached to the
bottom of the well, failure appeared to ke the
result of differential expansion between the
thermocouple wires and the surrounding Inconel
well, '

 
 

 

 

»

PERIOD ENDING JANUARY 31, 1958

“"'12.2, RADIATION DAMAGE *
G. _W. Keilholtz

tN PILE THERMAL CONVECTION LOOP TESTS

- ‘W. E. Brownmg -
H. E. Robertson - R.Ps Shlelds B

- The electrlcolly heated fuli-scoie mockup of the
m-pule thermol-convectlon loop, described previ-
ously,! was weided and filled with fuel in a dry
box.  The fuel. ‘mixture | BeF -LiF- UF, (37-62-1
mole %, fuel 130) containing’ depleted uromum and
normal lithium- WIII be circulated in’ the mockup,
which, like the m-plle loop, is fobrlcoted of
INOR-8 "alloy. Thermocouples - and heaters are
now being installed on the mockup to demonstrote
the workability of the design and ‘to" provide

operating . mformotlon for wuse in the m-plle loop |

tests.

Mechanical mockup tests were performed on a

series of thermocouple leads ‘being consrdered for

service in the coolmg-ctr annulus ‘of the loop. .

Since thermal expansion of the fuel tubes damaged

_ thermocouples in prevnously operoted Ioops, it was

desired to select a design which would prevent
plastic deformation of the thermocouple lead wires.

- The mechanlcol mockup test was performed in an
" appuratus which was similor to a portion ‘of the
“fuel ‘tube and coolmg-ulr-onnulus tube,,wnth the

outer parts mode of transparent plastic to permut

observotlon ‘of the thermocouple during expansion -

of -the fuel tube. The various thermocouple lead

wire ossembhes "being " consudered were tested .in -

" this: apparatus in order to determme ‘the. effect of \ for krypton after the heat treatment than it had

:'ff‘;longltudmol movement of the fuel ‘tube of as muchf‘;‘f\?{—-:- been before. - This effect was no doubt the result

- _r—r,os ,5 in. Radlol movement ot the fuel tube in the__i"": of dehy drot|on of the charcoal
- loop: is prevented by radial Spocer pins. . The

_“thermocouple " assembly selected - consists of - -'ftude of undesirable effects of the Li¢ isotope in

‘ "‘j:-;'_'r;"g"thln-wolled stomless steel tube which fits closely :'.‘fi';‘r.ithe fuel for the in-pile loop. The effects con-

"*",i,_'_'ioround ‘a’ ceramic’ msqutor tube. ~ The stomlessff-’

. “steel “tube-is. SPI" at the end and is. res:stance;;fi;; result trom obsorptlon of thermal neutrons by the

Lié, (2) ‘the brologrcol hozord that would be

  

-~':-:spot welded to the fuel: tube ‘near the point. where"-;"_;;"
= “the thermocouple bead is “joined to the fuel tube.: - present ‘in the event of an ‘accidental release of

" This - stainless - steel thermoc0uple |ocket is ’“'f"v-r*’-f'_?the tritium that would be produced by the (n,a)

. stalled neorly parolle! fo the fuel tube and- l"''-‘"-‘-':’-?_:.!recrc:t|on in Li%, (3) the internal ‘gas pressure that

- brought ‘out along @ straight line at o smoll angle ‘would be ‘generated by nuclear reaction products

"from the fuel tube.- As the fuel tube moves Iongl-r-‘-;é_:"*

 

lW. E. Browning et al., MSR Quar. Prog. Rep Oct. 31,
1957, ORNL- 2431 p 30.

   

 

tudinally during thermal expansion, the thermo-
couple jacket moves longitudinally along its own
axis through the port in the air annulus and through |
the heater assembly, and it moves slightly with
respect to the bundle of lead wires. This arrange-
ment provides a relatively rigid structural member
which does not move with respect to the thermo-
couple wire or the thermocouple bead, and thus

does not apply stress to the thermocouple. Thermo-

couples of this type have survived dozens of
cycles in which the thermal expansion was many
times greater than that expected in the loop.

The -in-pile loop is to include a charcoal trap
for removing the xenon which will diffuse from
the surface of the. fuel. = Any orgomc material
and water present in the charcoal as an impurity
would be decomposed by radiation from adsorbed

fission products, and the decomposition products
- could contaminate the fuel ond interfere with
interpretation of the results of the experiment.

The charcoal used in the loop will therefore be

baked in vacuum for 48 hr at 500°C in order to

decompose organic impurities. The charcoal which
was installed in the mockup loop was subjected
to this treatment. The holdup time of the heat-
treated. charcoal for radiokrypton in a flowing
helium stream was measured to determine whether

‘the heat treatment had reduced the capacity of the
- charcoal for noble gases. It was found that the
. charcoal was 10% more effective as an absorber

“Calculations were made to determrne the magm-.'

?-.SIdered ~were (1) the flux clepressron ‘that would ~

- from Lib, ~and - (4) ‘the possible chemical effects

of the tritium equrvolent of HF and the consequent
interference with the interpretation of the results
of the experiment. All these effects were found

79

 
 

 

 

MOLTEN-SAL_T REACTOR PROGRAM PROGRESS REPORT

to be unimportant except the last one. The

possible effects of traces of tritium in the loop
connot be evaluated with sufficient confidence
to risk operating the in-pile loop under such

conditions. [t was decided therefore to obtain

the best available Li’ for use in the in-pile test.
Lithium containing 99.8% Li’ was obtained and
was used in the production of a batch of fuel 130
(see Chap. 2.3 for production details)., Sufficient

fuel for filling three loops 'is now uvo:lable in

powder form

Preporations are being made in the ORR fo-, :

cilities for the operation of a similar forced-

circulation loop. The loop assembly housing and

lead tube have been fabricated for use in pre-
liminary hydraulic tests of the reactor before
startup. - These tests will show whether the lead
tube is structurally adequate in the reactor cooling-
water stream. The same assembly is being used
for surveys of ‘the flux in the experimental space
designated for this experiment., Conduit trays
and an auxiliary control panel are being installed
to carry lead wires from the experimental assembly
in the reactor to the laboratory, where the controls
for the loop will be located.

IN-PILE STATIC CAPSULE TESTS
H. L. Hemphill

Preparations for the irradiation of fuel 130 in
INOR-8 capsules in the MTR were continued. The
fuel used in these test capsules will be the same

w. E. Browning

as that used in the in-pile thermal-convection

loop. The INOR-8 tubing intended for use in
fabricating these capsules had to be rejected
because of high aluminum content. Another batch

80

of INOR-8 tubing of satisfactory composition con-
tained a large number of flaws. Short sections of
this batch of tubing are being selected for use

in the MTR capsules. The INOR-8 tubing for the

MTR capsules is more difficult to fabricate than
that used in the thermal-convection loop because
of the small diameter and large wall thickness,

- 0.200 in. OD ond 0.100 in. ID.

Other capsules are being fubricdted for irradi-
ation in the MTR in order to determine the stability

 of graphite in- contact with fuel 130 during ex-
posure to radiation. Existing Inconel capsules

are being used for the outer containers. Graphite

liners with 0.025-in.-thick walls have been pre-

pared that have a 0.100-in.-dia bore and are 1 in.
in length. These graphite liners will have o loose
plug at the top end and three small holes near the

~ bottom to allow the fuel to fill the space inside

the graphite liner and the space between the liner
and the Inconel outer can. The fuel will provide
heat transfer between the graphite and the Inconel
wall, and fairly accurate estimates of the graphite
temperatures can be based on measurements of
the temperature of the wall, the thermal con-
ductivity of the grophite, and the heat generation

by fission in the fuel. These'capsules will be

irradiated ot a metal wall temperature of 1250°F.
After irradiation the capsules will be examined

in the Solid State Division hot cells in order to

determine whether the graphite dlsmtegrated I

the graphite is intact, an attempt will be made to

determine the distribution of fuel components and
fission products in the graphite. The grophite
and the Inconel will be examined metallograph-
ically for evidences of corrosion, and the fuel will

be chemically analyzed for carbon and the con-

stituents of Inconel.

 
o

 

PERIOD ENDING JANUARY 31, 1958

2.3, CHEMISTRY

W, R. Grime_s_ o

FHASE EQUILIBRIUM STUDIES .Y
Systems Contoming UF and/or Tl\F

R. E Thomo H. A. Frledmon
H. lnsley C. F. Weaver -~

Phase eqmllbrlum studles are belng mode for
‘determining - whether an LiF- Ber- ‘mixture will
dissolve ‘sufficient - ThF, and" UF, to provide a

“fuel for a fused-solt breeder reactor.
- practical to treat the quoternory system LiF- Ber-
ThF ~UF, as o fernary system if the “phase

--equnlrbrlo and isotherms ore’ very SImllOl’ in the
‘and LiF-BeF -UF ‘and

systems LiF-BeF -ThF
it exfensive solld2 solutlon occurs - in

 LiF-ThF -UF o
" The systems BeF -ThF (ref 1) oncl BeF -UF

~are very similar, quuadus curves of ‘the systems
LiF- UF4 (ref-2) and LlF-ThF‘(ref 1) are also
similar, although - LrF-UF and LiF-ThF“, ‘solid
- phases are, in generol d:ssnmllor.
‘that phase diagrams of the systems LiF-BeF -UF
“and LiF- Ber-ThF

tlte system

wrll

LiF-ThF -UF
mterpolotrons can be made between the systems
LiF-BeF -UF ¢ ond LiF- BeF -ThF, ‘with regard

to llqmdus temperotures and phase relotlonshtps. _
The System LrF-BeFZ-ThF . -

- “system LiF- Ber-ThF o
“"Z',";luqu:dus proflles obtamed prewously.

 

Recen '

R "}i;'.',:studres

  

    
   
   
      
    
   
  
  

  
 

  

-,.";Tclaoorom shown ine Flg. 2.3.1.

 
 

 
  

,;thermal-grod nent quenchmg expenments. o

 

 

o }1957. ORNL-2431, p 36. -

'(1958)

 

“concentration

It will be

It is eVrclent .

be similar ond thot'
- substitutional solid solution ‘occurs (in the system -
It - may be inferred that some

Results of
phase - equ:llbnum studres now, under way ot the -

 

 

experiments k__hove been confmed princnpally to.
/ -of “the “section - of “the - “diagram ‘that
":represents mlxtures contomlng 60 to 80 moler%_" '

 

B Tl'ns d‘ogrom 15'-__
- :'based ‘on (1) onolyses of the thermal - effects that. -,
o are’ indicdtedin heating ‘and coolmg (from’ above

- ":‘the llqmdus)'f‘curves tor 50- o IOO-g fused-solt

r.exommotrons of these somples, ond (3) reSUlts offf

2(:. J. Barton et al.. 5 'Am. Ceram. Soc. 41 63—69 o
' twln_ne_cl and biaxial rnegotwe. :

: Br_eecler reactor blanket or breeder reactor fuel

solvent ~composifions, whose maximum ThF,
is restricted to . that available in -
salts having less than o 550°C liquidus, may be

-.chosen from an area of the phase diagram (Fig.2.3.1)

in which the upper limits of ThF 4 concentration

'ore obtolned in the composmons

75 mole % LiF—16 mole % ThF =9 mole % BeF,
9.5 mole % LiF~21 mole % ThF ;~9.5 mole % BeF,
__;s'a*;n;,;, % LiF~22 mole % ThF;~10 mole % BeF

Results of thermol-grodient quenchlng experiments

_'__~(Toble 2.3.1)} have shown that the temperatures
“along two of the boundary curves limiting the

prrmory phase 3LiF:ThF, are those shown in
Figs. 2.3.2 and 2.3.3. lte results of quenched
samples 2, 4, 5, 6, and 7 show that a lower melting
liquid occurs in the system than had been suggested
by eorlrer data, and yet neither petrographic nor

. x-ray diffraction data permit a conclusion as to
* ‘what phases are present in the eutectic which this
-low-meltmg liquid represents. The composition of

~ this liquid appears to be approximately 66 mole %
-LiF=4.5 mole % ThF -29.5 mole % BeF,; how-

, 'ever, the sequence of appearance of solid phases
in “Table 2.3.1 does not suggest that this is @
*eutechc composition. '

1 The System LiF-ThF -UF . - The compounds

- ThF, and UF‘ form a contmuous solid solution.
conslstent with” 7 &

\";Slowly
i::breoks, ‘most-of which are at lower temperatures
than - ‘the - lnquiclus values: ‘observed - optically .in
‘f"}uenched samples, . o
he " slowly cooled: melts, the “optical results from
'uenched somples are relned on for llqu:dus volues.' e

cooled ‘melts’ give . multlple temperature

 
 
  
  
   
 
 

 

“Since- supercoollng occurs in

’]"quurclus -points and ‘curve are: shown in
igs 2.3.4. - Apparently no minimum’ exists within
he system; consequently, the’ ‘solidus - line: connot‘
xtend: ‘below the melting. point - of UF,. The

lidys line must therefore be close to the‘lliqmdus

  

 

;'Ime however, the- presence of a brown, unidentified
“mdterialat temperatures. below ‘the llqmdus has
--J-__precludecl optrcol observot:on of soltclus temper- '

C J. Borton et at., MSR Quar. Prog. Rep. Or:! ‘31.’_- _:..,V - utures- ol

The ThF4 UF4 solad soluhon is green, frequently
It has an optic

81

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

angle 'of about 65 deg aond a birefringence of

‘approximately 0.05. The indices of refraction are

shown in Fig. 2.3.5 as functions of UF content.
The compounds LiF4ThF, and LiF. 4U|=4 form
a continuous solid solution. A tentative liquidus
curve for the LiF. 4ThF ,-LiF. -4UF , join is shown
in Fig. 2.3.6. This liquidus ond the liquidus in

the system ThF 4-UF ; indicate that the ThF +~UF
' sohd-soluhon prlmary-phase field is shghtly

7 LiF-6 ThF,

45
40

35
o0
R .
° ——
£ P-595
»

&Y E-580

3 LIFThF,

E-570

20

10 < 700

750

* convex in the direction of decreasing temperatures.

The three-phase region composed of liquid,
ThF ,-UF  solid solution, and LiF- 4ThF4-L|F -4UF ,
soh soluhon, which emerges as the LiF. 4ThF4-
L|F-4UF4 solid ‘solution primary-phase field, is
convex in the direction of increasing temperatures.

The LiF. 4ThF4 -LiF-4UF, solid solution is

green and biaxial negative, ond has an optic angle

of about 5 deg. lts birefringence is approximately -

. UNCLASSIFIED
ORNL-LR~DWG 27910

| LiF-2ThF,
T E == EUTECTIC
p = PERITECTIC .
. LIQUIDUS
TEMPERATURES
ARE IN °C

 

 

 

 

 

_ ‘ 35 40 a5
‘2L|F-ae|=2/ - 5\P-490 ' :

BeF, (mole %)

i;'ig; 2.3.1. Parhul Phase Dmgrom of the System LiF- Ber-ThF for ihe Region Represenhng Mnxfures Con-

taining 50 to '|00 Mole % L|F

82

 
 

PERIOD ENDING JANUARY 31, 1958

'TGbI_Qé; 2.3.'_17_; :R.e'sdlf's of 'Thermnlgétudlent Qu'enching Experiments on the System Lil”-Ber--'I"hF4

 

Composition (mole %) " " Temperature

 

 

7 Sampie B — : g Phases Present
No. LiF .BeF,  ThF, Q)
1 68 27 5 460 3LiF-ThF * and liquid
| 448  3LiF-ThF , 2LiF-BeF,, and liquid
426 3LiF-ThF4, 2LiF-BeF2, and little or no liquid
2 6 2% 5 452 3LiF-ThF, and liquid
' ' 444 3LiF-ThF ,, 2LiF-BeF,, and liquid
413 3LiF-ThF ,, 2LiF-BeF,, ond very little liquid**
3 e 3 5. 451 3LiF-ThF,, and liquid
- 431 . 3Li_F-ThF‘, 2LiF-BeF2, and very little liquid
428 3LiF-ThE4,'2LiF°BeF2, and no liquid
4 65 29 .6 466 TLiF-6ThF and liquid
| | 457 3LiF-ThF, and liquid, no 7LiF-6ThF,
-440 - 3LiF-ThF,, 2LiF-BeF,, and liquid
'4_1_37 - 3LtiThF4, 2LiF-BeF2,'and very little liquid
5 67 . 26 - 7 470 3LiF-ThF, and liquid
' | 444 3LiF- ThF,, 2LiF-BeF,, and liquid
| 413 3LiF-ThF,, 2LiF-BeF,, and a small amount of liquid
6 6 - 21 7 474 ~ JLiF-6ThF, and liquid
o 457 3LiF<ThF, a small amount of 7LiF+6ThF ,, and liquid
452 3LiF-ThF, ord liquid, no 7LiF-6ThF, |
444 LiF ThF,, 2LiF-BeFy, and liquid

o413 'rftf;rsl_.F.'th 2L1F-BeF2, ancl ||qund o

 
  
  
  
 
  
  
  
  
 

479 7L|F 6ThF and hquud ‘ :
L 45S  TLIFe 4ThF , 3LiF-ThF,, and l-qwd
16 - ".-':;'SLIF-ThF 2L|F-BeF2, cnd ||qutd T ,
‘3L:F-ThF _2LrF-BeF2, and llflle or no ||qu1d .
""""‘ffaL.F-'th and liquid -~
l-'-.-:f;;73LiF‘ThF and |iqund
'*’:j}fl_,F 6ThF, und hquud
.77_.3L|F-T|1F 7L1F-6ThF 4 ond hquud .
B :3LFF "th 7L;F 6TI1F4, ancl a smull amount of hqund

 

7 _ , *The phuse 3LiF. ThF4, as l‘l‘ appears in the fernory system, is routinely observed as a solid solution.
; v - o **The presence of liquid in: the Iowest femperuture ‘sample indicates that the temperature range of the experiment
T was not iow enough to permit determinutlon of fhe solldus temperufure. ' C

83

   

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED -

. ORNL—LR—DWG 27944
560

 

 

540 '
o B N
520 \

500

480 \
460 |—— — -
440

420 — _ \

400

 

 

 

 

TEMPERATURE (°C)

 

 

 

 

 

 

 

 

 

 

30 25 20 15 10 5 0
ThF, (mole %)

Fig. 2.3.2 Temperoture Variation Along the Boundary
Curve from the Composition Point 71 Mole % LiF-29
Mole % lThF‘ to the Eutectic Composition Point at
About 29,5 Mole % BeF2—4.5 Mole % ThF  -66 Mole %
LiF.

UNCLASSIFIED
ORNL-LR-DWG 27943

]
Q
a

   

- LIQUID

3

LIQUID + ThF, - UF, SOLID SOLUTION

TEMPERATURE (°C)

200 ThF;-UE. SOLID SOLUTION
BOO
Thi, - 10 20 30 40 50 60 70 BC 90 U
UF4 {mole %} 7

Fig. 2..3._4. The System ThF‘-UFf

0.01. The indices of refraction as functions of
~ UF, content are shown in Fig. 2.3.7.
~ The compounds 7LiF6ThF, ond 7LiF.6UF,
form a continvous solid solution. The liquidus
curve for  the 7Li’F'6T|'1F4--7‘|..iF06UF4 join is’
shown in Fig. 2.3.8. This curve and the three-
" phase region in the LiF-4ThF ,-LiF-4UF , join
indicate that - the LiF-4ThF4-LiF-4UF4 solid-

“solution . -primary-phase field is convex -in the. .

direction of increasing temperatures.

UNCLASSIFIED

' ORNL~LR—DWG 27942
580 ,

560 \\
540 | \
ol 1)
TN
NEEERAV

440

 

 

 

 

 

 

TEMPERATURE (°C)

 

 

 

 

 

 

 

 

 

420

 

30 25 20 15 10 5 0
Thi, {mole A

Fig. 2.3.3. Temperature Variation Along the Boundary
Cutves from the Composition Point 78 Mole % LiF-22
Mole % ThF, 1o the Composition Point 63 Mole %
LiF~37 Mole % BeF,.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
, ORNL—LR~DWG 27944
§.62
1.60 —
"]
]
g 1.58 =
P /
g 7
156 - —
E / ' ' |1
"1 ; /l»
> b
8 152 L~
= /'/‘I
150 e
1.48 - -
ThE, 20 40 60 80  UR
: UR {mole %)
Fig. . 23.5 Indices of Refraction in the System

The 7LiF-6ThF -7LiF6UF, solid solution is

green and uniaxial. lts birefringence is low-and

_varies from 0 to 0.006. The indices of refraction
~ are ~shown .in Fig. 2.3.9 as functions of UF,

 
 

 

 

 

PERIOD ENDING JANUARY 31, 1958

: : _ ~ UNCLASSIFIED ' _ UNCLASSIFIED
: . - ORNL-LR-DWG 27915 . ORNL-LR-DWG 27916
1050 1.62

' .
1000 \'\\ LIQUID 1.60
\

950 e —

 

 

 

 

 

 

 

\ _ 1.58
LIQUID + Thfy~ UF SOLID SOLUTION | \.\

EQ'N‘

.'\

LIQUID + ThFy—UF, SOLID" 7k" .
y

800 I"SOLUTION + LiF - 4ThF, =

T
LiF- 4UF, SOLID SOLUTION—] ‘ =~

750 . - 50
LiF-4ThF4—LiF-4JUF4 SOLIID SOLUTION -

./ /

 

 

9200 1.56 | "]
) /l

o

 

 

 

 

INDEX OF REFRACTION

\\

1.54

 

TEMPERATURE (°C)

 

 

1.52

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o - 7 0 20 30 40 50 60 70 80
UF, (mole %)

 

 

 

 

 

700

 

0 10 20 30 40 50 60 70 80
UF, (mole %)

Fig. 23.7. Indices of Refraction Along the Join

Fig. 23.6. The Join LIF-4ThF -LiF-4UF . LiF-4ThF -LiF-4UF .

UNCLASSIFIED
ORNL-LR-DWG 27917

 

850

 

800 -
LIQUID

~. | | T

750 -

 

*— '
NN ) |
NN - LIQUID + LiF: 4ThF,~ LiF - 4UF, q
TO00 p————S—PN———— SOLID'SOLUTION = —p—
] LIQUID+LIF-2ThE, N o L ,
- SOLID SOLUTION * Lk | S

 

 

 

 

 

 

- LiF :2ThF, SOLID SOLUTION 1 -

TEMPERATURE (°C)

650 —L— LIQUID + LiF - 4Th, ~LiF - 4UF, SOLID SOLUTION +——

= -———— —————————f_—_.

 

‘—-—-7'

 

600 fu

 
 

| LIQUID + LiF -2Thf, SOLID SOLUTION + Y- | =

1{Lic - -2 Thi N T '_-"-Luo’u:_o +LiF- 4ThF, ~ LiF - 4UF, SOLID SOLUTION | -
7LIF+6ThF, — 7LiF - 6UF, SOLID SOLUTION/- *{ = |~ £ 7LiF - 6ThF, - 7LiF - 6UF, SOLID SOLUTION | -

 

. TLIF-6THF,“7LIF 6UF, SOLID SOLUTION

 

 

 

 

 

 

 

 

 

 

 

500

 

0 5 .- A0 15 ZQ : 25 30 35 40 45
UF, (mole %)
Fig. 2.3.8, The Join 7LiF-6ThF,-7LiF-6UF,.
. L 85

 
 

 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED
ORNL-LR-DWG 27948

 

1.62

 

1.60

 

1.58

 

1.56

\

 

1.54 ' o

/
4/

INDEX OF REFRACTION

 

152 2
1.50

 

 

 

 

 

 

 

 

 

 

 

 

0 5 410 5 20 25 30 35 40 45
UF, (mole %)

Fige 23.9. Indices of Refraction Along the Join
7LiF'6ThF4-7LiF-6UF4‘.

ThF,

content. As the UF, content increases, the solid

solution changes from uniaxial positive to uniaxial
negative. Further, for mixtures containing about
31 mole % UF4, the solid solution should be
isotropic. Quenches of such mixtures have
produced 7LiF.6ThF 7LiF.6UF, solid solutions
that have birefringences which are too low to
measure microscopically.

The curvatures of the ThF -UF , solid-solution
primary-phase field ond the L‘iF-41'hF4-LiF-4UF4
solid-solution primary-phase field, which were
mentioned above, indicated that their boundary
curve should bend away from the LiF vertex of
the LiF-ThF‘-UF4 diagram, Experimehtal values
confirm this indication, as shown in Fig. 2.3.10.

UNCLASSIFIED .
ORNL~LR—DWG 27949

 

 

 

a0
80
70
/\/ ThF,- UF, A
SOLID SOLUTION
6\; 60 3
0\0 - .
£
Qb
& 80
40

 

 
 
  
 
  

LiF - 2ThE,
SOLID SOLUTION

3Lif - ThF,
. SOLID SOLUTION

LiF~4ThF4—LiF-4UF‘w/\/\
SOLID SOLUTION

  

\V
/N

 

LiF & ' -
{0 20 \ 20 40 50
4LIF- UF,

&

 

60 70

UF, (mole %)

Fig. 2.3.10. The Primary-Phase Fields in the System LiF-ThF4-UF4.

86

 
 

 

 

\J

In the system diagrom presented in Fig. 2.3.11,
the temperatures of points a and ¢ are 500 and
488°C, respectively. Consequently the LiF-
4LiF.UF, boundory curve temperatures drop from
point a to point c. The temperatures of points ¢
and 4 are 488 and :500°C, respectively. Conse-
quently the boundary curve between the LiF
primary-phase field and the 7Li F-6ThF ,-7LiF.6UF

solid-solution primary-phase field drops in temper--
.ature from point d to point ¢. In addition, evidence

exists (Table 2.3.2) that 4LiF.UF , appears as a
third solid phase in compositions along this

PERIOD ENDING JANUARY 31, 1958

boundary curve. There is also evidence (Table
2.,3.2) that LiF appears as a third solid phase in
compositions along the boundary curve between the
4LiF-UF , primary-phase field and the 7LiF 6 ThF,-
7LiF-6UF ; solid-solution primary-phase field.
Thus the boundary curve temperatures must
decrease from point & to point c. Point c is, then,
a eutfectic.

The solid phases in equilibrium in compositions
near point ¢ at the invariant temperature are LiF,
4LiF-UF,, and 7LiF-6ThF -7LiF.6UF ; solid

solution containing about 41.5 mole % UF ,. From

UNCLASSIFIED
ORN{—-LR—-DWG 27920

 

LiF-2ThF,
SOLID SOLUTION

 

  

LiF-4ThF, - LiF-4UF,

 

 

 

SOLID SOLUTION

  

 

 

 
 
  

\ ' 7LiF-6ThE,—
N 7LIF-6UFR,

 

 

 

'UI-;," (mole %)

Y SOLID'SOLUTION

  

 

   

  

  
  
 

-7 \/ L _ . |
« 30 . 35 40 45 50
T 4LIF - UF, ' o R

Fig. 2.3.11. Primary-Phase Fields in the .System LiF-ThF4-UF4 for the Region Representing Mixtures Con-

taining 50 to 100 Mole % LiF.

87

 
 

 

 

 

MOLTEN.SALT REACTOR PROGRAM PROGRESS REPORT

- Table 2.3.2. Results of Thermal-Gradient Quenching Experiments in the System LiFf_ThF“-UF‘

 

Composition Temperature of

(mole %) : Phase Change

Phusés Above Phase Change

Phases Below Phase Change ‘

 

7LiF-6ThF4 sqiid solution

1 e . . : o
LiF | ThF4 l.lF4 ("C)
40 10 50 808 Liquid Liquid + LiF+4UF ,«LiF-4ThF , solid
- o - solution
33.3 33;3 33;3 862 Liquid + UF4-ThF4 solid solution 'Liquid + LiF'4UF4-LiF°4ThF4 solid
' . o sofution -
40 50 10 880 Liquid Liquid + LiF°4UF4-Li F+4ThF, solid
: solution
725 1 26.5 485 Liquid + 4LIF-UF, + 7LiF+6UF ;- LiF +4LiF-UF , + 7LiF-6UF -7LiF+6ThF ,
C 7LiF°6ThF4 solid solution solid solution
71 2 7 27 491 Liquid + LiF + 7LiF-6UF4- LiF ~l-~4Li-F-UF4 + 7LiF'6UF4-7LiF-6ThF4

solid solution

 

this information the compatibility triangle at the
invariant temperafure can be drawn as shown in
Fig.2.3.11. The invariant falls within this triangle,
and consequently, as stated above, point ¢ must
be a eutectic.

Solubility and Stability of PuF,
' Molten Fluorides

C. J. Barton R. A. Strehlow

Values for the solubility of PuF, in NaF-BeF,
(57-43 mole %) at three temperatures were given in
the previous report.d During the past quarter,
values were obtained for the solubility of PuF
two additional NaF-BeF mixtures, three LiF- BeF2

composmons, and one NaF- LlF -BeF, mixture

ds a function of temperature. The range of BeF,
concentrations covered in each of the binary

solvent systems, approximately 36 to 50 mole %,

includes _the concentration range likely to be of
interest in the MSR program. Binary mixtures with

lower BeF, concentrations have too high liquidus -
B temperatures to be useful at power reactor temper-
atures, and the viscosity of mixtures containing

more than 50 mole % BeF, is too high to make
them attractive. The ternary solvent mixture was
included in the ‘study to permit solubility de-
terminations with a lower BeF, concentration than
would have been possible in the binary systems
at the lower limit of the temperature range studied.

3¢, J. Barton et al,, MSR Quar. Prog. Rep. Oct. 31,
1957, ORNL-2431, p 39.

 

88

The results obtained to date in these studies
are presented in Table 2.3.3. Some of the results
are almost certainly incorrect and will be checked
in new filtration experiments ‘as soon as possible.
The possibility of analytical error was minimized
in most of the doubtful results by repeating the
analysis, either with a new portion of the same
sample or a separate portion of the origineal
sample., Except for the first few samples sub-
mitted for analysis, which were ground but not
sieved, all filtrates were ground either to ~65 or
~200 mesh and mixed thoroughly to avoid the
possibility of inhomogeneity. A possible cause of
the observed discrepancies in the data would be
an error in the measurement of the temperature of

-the liquid salt mixture. A fairly large vertical

temperature gradient is known to exist in the
heated zone occupied by the fused salt container,
and it is possible that the thermocouple was
accidentally shifted in the course of the experi-

‘ments, An additional uncertainty in the determi-

nation of the melt temperatures resulted from the
thermocouple wells in most of the filter bottles
used in these experiments being shorter than the
deéigh"'cu“ed for; the ends of the wells were not
in contact with the melts. Improved filter bottles
that have thinner walled thermocouple wells of
the proper length to permit the " thermocouple
junction to be surrounded by the salt mixture
became available near the end of the quarter and
were used for ‘the last two measurements made,

 
 

 

. IJ'

PERIOD ENDING JANUARY 31, 1958

Table 2.3.3. Solubility of PuFy in Alkali Fluoride—Beryllium Fluoride Mixtures

 

Solvent Composition -

 

Filtrate Analysis

 

 

| Filfrdfien
(mole %) Temperature: Pu l:'uF3
NaF LiF BeF, (°c) (wt %) (mole %)
64 36 550 .54 0.29
598 2.43 0.46
650 4.40 0.85
57 43 538 1.17 022
' 600 1.36 0,26
652 2.16 0.41
50 _ | 50 552 1.77 0.34
' 600 1.97 0.37
651 272 0.52
64 36 532 1.15 0.16
' 600 1,82 0:27
643 4.30 0.63
56 44 550 | 1.98 0.30
' ' 649 6424 | 0,98
5146 48.4 463 1.02 0.16 -
| 549  2.44 0.38
599 2.88 0.45
654 5.76 0,93
56 16 28

554 7468 1.4

 

~ mole % LiF-28 mole % Ber

" nmproved boflles wlll be used to check some’ of thef:_f:'fi
The solvent composmons given in
The -actual compo-- . -

“sifions’ have ‘not yet been estabhshed by chemical -
~analysis, except for ‘the 51.6 mole- % LiF-48.4
composmon,whlch is_the some
mixture as that used for the determmuhon of the
_which _is. dlscussed m a..ieii'

-~ mole % BeF

a solubnln‘y of CeF

“subsequent secflon of this chapter. S
- The data _given in ‘Table 2,3.3 show thut for"

LiF- BeF2 '

that is, the measurements on the 51.6 mole %
LLiF-48.4 mole % BeF and 56 mole % NaF 17

earlier data.
Table 2.3.3 are theoretlccll

B mixtures

solubility of PuF is - hlgher in the mixtures with
50 mole % BeF qnd 36 mole % BeF than in the
mixture with 43 mole % BeF Thus there is. an

mdlccmon that the solubility of F’uF
. ‘solvent goes through a minimum in the wcm:ty of
“mixtures ‘with 43 mole % BeF.,.
,_‘vafue obtamed with the ternary mixture, which is
-~ probably low because of the lack of a sufficient
~amount of PuF, to saturate the solutlon, mdlcates

mlxtures. _ The -

'.‘wnh decreasmg BeF
plutonium in thls mlxture wasfound fo be combmed .

-f‘os NaPuF 2

in the composmon ranger'-»
"studled ‘the.. so]ublllty of l'-’uF3 increases with -
_increasing - BeF2 concentration, “at . least at ‘the
lower temperatures. - In. the NaF- Ber system, “the

in this

‘The solubility

that the’ solubfifly of PuF continves to. increase
concentrcmon. All the

“The - solubtllty of . PuF “at’ 565°C_

in_the blnary mixtures studied varied’ from about
- 0.2 mole% for the 57 mole % NaF ~43 mole % BeF,

 mixture ~to 0,45 mole % for the -51.6 mole %_
-_!_fLiF--48.4 ‘mole % Ber mixture.
tration. range is believed, at the present time, to -
“be: more .than - adequate to fuei a molten: salt
-fplutomum-burner reactors 2

Th:s concen-

Another purpose of this investigation was the
observation of the stability of PuF, in fused salt

89

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

mixtures. This has been accomplished by de-
termining the plutonium species present in cooled
samples of filtrates and unfiltered residues
produced by filtering mixtures that had been
heated at temperatures of 550 to 650°C for periods
that were usually in excess of 2 hr. The plutonium
compounds present in the mixtures were identified
by petrographic examination, and valence de-
terminations were made by spectrophotometric
examination of dissolved portions of the samples.
Only trivalent plutonium has been found in the
mixtures examined to date. In addition, the bottom
portions of two nickel filter bottles used in
~ filtration experiments were submitted for analysis
after removal of all except a trace of fused salt.
 One sample had been in contact with the NaF-
BeF ,-PuF, mixture ot 600°C for about 2 hr and
the other had contained the LiF-BeF ,-PuF

“mixture at temperatures varying from 500 to 650°8

for approximately 8 hr. The plutonium found in the

two samples amounted to 0.05 and 0.17 mg,
respectively, and could be accounted for by the
trace of fused salt remaining on the nickel. No
evidence of disproportionation of PuF, in fused
alkali fluoride-beryllium fluoride melts under the
conditions maintained in these experiments has
been observed to date.

FUSED CHLORIDES AS SECONDARY HEAT
TRANSFER FLUIDS

R. E. Moore

The physical, chemical, and nuclear properties
of a secondary coolant for a molten.salt power
reactor must satisfy a number of requirements.
The following characteristics are desired:

1. low melting point,

2. low viscosity,

3. high heat capacity,

4. high thermal conductivity,

5. low vapor pressure,

6. stability toward structural metals,

7. -reasonably low thermal-neutron cross section,
8. freedom from activation in a radiation field,

_Fluorldes and chlorides seem to offer the best
~choices among possible salt systems. Melting
points of 300°C and lower can be obtained with
fluoride mixtures containing ‘BeF,, but the
viscosities of these mixtures are hngh. If a low
melting point accompanied by a low viscosity is
to be obtained it is necessary to turn to chloride
systems, )

90

- A large number of chloride systems have been
described in the literature, but, after an elimination
process - based on gross deviations from the
properties listed above, there remained only a few
systems for further consideration. The eutectic
compositions (mole %) and melting temperatures of
these systems are listed below:

Melt.ing

VComposiiion_ Temperature

(°c)
41.7 mole % RbC|-58.3 mole % LiCl 318

© 64 mole % ZnC|2-36 mole % SnCI2 : 171 -
38 mole % KCl—62 mole % SnCl, 180
29 mole % KCI-71 mole % ZflCl2 262
23 mole % LiCl~77 mole % ZnCl, 294
47.5 mole % RbCl-52.5 mole % ZnCl, 249

The very low melting points of the last five
mixtures listed would seem to make these compo-
sitions especially attractive. All these mixtures
contain, however, either .‘:'mCl2 or ZnCl,, which
have high vapor pressures, although the pressures

are possibly not high enough to interfere with

pump operation in the coolant circuit. A calcu-
lation of the ideal vapor pressure of the 29 mole %
KCI-71 mole % ZnCl, mixture, based on literature
values for ZnCl,, gave values of 140 mm Hg at

650°C and 1.4 mm Hg at 450°C. The actual

pressures may be somewhat lower than the ideal.
Yapor pressures of the mixtures containing SnCl,

would be expected to be much higher than those'

for mixtures containing ZnCI2 because the boiling
point of SnCl (623°C) is considerably lower than
the boiling pomf of ZnCl (732°C) -

Another 1mportant questlon concerning compo-
sitions containing SnCl, or ZnCl, is the possibility
of reaction with structural metals. Calculations
based on vqlues of free energies of formation?
indicate that considerable reaction may occur,
especially in the case o{:'SnCI2 mixfures. There
is enough uncertainty in the calculdtions, however,
that static corrosion tests should be made. Plans
are being made for experiments in_which Inconel
will be exposed to molten mixtures containing

ZnCl, and SnCl, in seoled containers of fused

silica.

 

4A. Glassner, A Survey of the Free Energies of

Formation of the Fluondes, Chlorides, and Oxides of

the Elements to 2500°K, ANL.-5107 (Aug. 1953)

 
1) ,!f\ N

 

 

 

The eutectic composition41.7 mole % RbCI-58.3
mole % LiCl appears to be the most attractive
from the standpoints of vapor pressure and
corrosion. The boiling points of RbCl and LiCl
(1390 and 1353°C) are sufficiently high that the
vapor pressure of the mixture should be negligible
at reactor temperatures. Calculations based on
values of the free energies of formation? indicate
that the mixture should be extremely stable in
contact with Inconel. Thermal-convection loop
tests should be made to determine the rate of
mass transfer. The new low prices of rubidium
salts (quoted by American Potash & Chemical
Corp. at $13.00 to $27.50 per pound) may remove
the cost objection to the use of rubidium chloride.

The treatment of RbCI-LiCl mixtures to remove
water, which is always present in the salts,
requires vacuum drying, grinding, and contacting

with HCl during slow heating. Once the hydrolysis

reaction
Cl= + H,O==0H" + HCI

has proceeded to the right, the reaction is not
readily shifted to the left.®> An apparatus for
carrying out such treatments is to be constructed.

VAPOR PRESSURES OF LiF-BeF, MIXTURES

F. F. Blankenship
It was anticipated that at MSR operating temper-

S. Cantor

atures the vapor pressure of the fuel mixture would -

be low, but, since it is important to know the
magnitude of the vapor pressure, the total vapor

-pressure of a fuel solvent composed of 64.9 mole %
LiF and 35.1 mole % BeF, was measured by using
 the Rodebush Dixon® method. The equcmon thot
Hits the dato, whlch are gwen in Toble 2.3.4, is.

1 0,050

logp(mmHg)---rsm o

.‘.there T is temperotu:e in’ °K, A Ilneor extropo-‘
~ lation of the data to temperatures of reoctor mferest i
is presented in Tobfe 23.5. Ik ,
From’ the - vapor. pressures of the NoF BeF ; Lo
system,? it was found that a change of 5 mole % .
. m the h|gher bodmg component lowered the 1otal,"

 

5H A, Lumnen, W S. Ferguson, ond R. A, Osteryoung, '

J. Electrochem. Soc. 104, 516~20 (1957).

SW. H. Rodebush and A. L. Dixon, Phys. Rev. 26,
851 (1925), .

PERIOD ENDING JANUARY 31, 1958

Table 2.3.4. Vapor Pressures of the 64.9 Mole %
LiF.35.1 Mole % BeF, Mixture

 

 

~ Temperature Pressure

(°C) 4 (mm Hg)
96846 4.9
1017 ‘ 9.4
1039 : 12.9
1066 18.1
1095 25.8
- 1128 40.0
1146 47.3
1160 55.4
1182 71.8
1194 83.0
1203 88.8

 

Table 2.3.5. Extropo_loted VYapor Pressures
of the 64.9 Mole % LiF~35.1 Mole % Ber Mixture

 

 

Temperature Pressure
(°C) (mm Hg)
500 0.000058
550 0.00036
600 0.0018
650 0.0076
700 0.027

 

pressure by approximately one-half.  Similar
behavior can be expected in the LiF-BeF, system.
For instance, the vapor pressure of a 70 mole %

LiF-30 mole % BeF, soclution would be about
~one-half the vapor pressure’ of -the 64.9 mole %
LiF=35.1 mole % Ber soluhon ot fhe same.
temperature. ' S ' S

FUEL REPROCESSING _
G.M Wofson - F.F. Blankenshlp

Solubilnty of Noble Goses in Molten Fluorlde
' M:xtures P :

"NV, szth

Numerlcol values of the solublhhes of hellum,

“neon, argon, and xenon in NaF-ZrF (53-47 mole %),
- expressed as Henry s Iow constonts, were presented

 

7K A, Sense, R. W, Sfone, and R. B. Fiibert, Jr.,
Vapor Pressure and Equilibrium Studies of the Sodium
f‘éu;;zde-—Beryllzum Fluoride System, BMI-1186 (May 27,
5

AN

 
 

 

 

- previously.8  Measurements on the solubility of -
helium and of neon in NaF-KF-LiF (11.5-42-46.5 -
mole %) have now been concluded, and the experi- -
mental results are summarized in Tables 2.3.6 and
2.3.7 and are presented graphically in Figs. 2.3.12,
2.3.13, and 2.3.14. The solubility study was

undertaken with the NoF-KF-LiF mixture as the
solvent pending the completion of facilities for
studying solvents containing BeF .

The solvent NaF-KF-LiF has a liquid Structufé

which is quite different from that of the’ NaF-ZrF
mixture, and the structures of mixtures contcm:ng

BeF , probably range in between. Accordingly, it
may be safe to expect that the numencal values

 

8J. H. Shaffer et al., MS'R Quar, Prog. Rep Oct. 31,

1957, ORNL 2431, p 47,

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPO_RT

of the noble gas solubilities in solvents con-

taining BeF will ‘be less than the corresponding
values . f\lcF ZrF ‘but more than those in
NaF-KF- L|F. , ‘

The results presented in Tubles 2.3.6 and 2.3.7

show the same trends as those prev:ously observed
for solubilities in mixtures containing ZrF The

solubilities follow Henry's law, mcrease with

- increasing - temperature, and decrease with in-
. . creasing molecular weight of the gos. The heats
“of solution for the helium and neon gases in this

solvent were calculated to be 8000 and 8900

_calories  per gram-mole of gas. The numerical
_magnitudes ‘of the solubilities in NaF-KF-LiF ‘are
-roughly” 50%  of the corresponding values in
~ NaF-ZrF . ~ A more detailed analysis will be made
when the measurements ‘of the solublhty of argon
in thls solvent are completed

Taoble 2.3.6: Solubility of Helium in NaF-KF-LiF ('lslo5442-46-'5 Moie'%)

 

 

 

 

. Saturating S
Temperature Pressure Solubility K*
(°C) (atm) -+ (moles of helium/cm® of melt) ‘
x 10~8 x 10-8
600 2.08 21 106
1,77 190 - 10.7
1.51 16,9 - ' 1.2
1.00 1.0 | 11,0
1.00 12.8 - 128
AV 13t 07
650 2,08 285 o 13g
2.04 34.8 - A
1.75 30.8 178
0.98 176 ' 1749
| | | AV 17,5 1 0.2
L S 2.06 483 235
| 2.04 48.2 ) 23.6
1.77 a8 - 2346
1.51 336 o 22.3
0.99 S 217 21

 

Av 230 £ 0.7

 

*K = c/p in moles of gas per cubic centimeter of melt per atmospheres

wi

 
 

 

 

2}

"

PERIOD ENDING JANUARY 31, 1958

Table 2.3.7. S&I_dbflity of Neon in NaF-KFaLiF (11.5-42.46.5 Mole %)

 

" Saturating s'| .
olubility

 

Temperature Pressure K*
_ (°C) (atm) (moles of neon/em’ of melt)
x 10~8 x 10~8
600 207 9.57 4.61
1.49 6.42 4,30
1.01 4.20 4.16
' Av 436 t 0.20
700 2.05 15.06 7:36
' 1451 11.04 7.33
1.02 7.99 7.84
Av 7.51 t 0.22
800 : 2.07 2252 10.89
1.50 16+53 11.00

1.03 12.05 11.66

——————n.

Av 11.18 * 0.26

 

- SOLUBILITY (moles of He/em of meit)

*K = ¢/p in moles of gas per cubic centimeter of melt per atmosphere,

 

 

 

 

 

 

_ UNCLASSIFIED
16 UNCLASSIFIED (x10~7 ORNL—LR—DWG 27922
0 ORNL—LR--DWG 27921 4.0
6.0
S$ y 3.0
S ’ oG
5.0 «°
/ ° G
109

4.0

 

 

 

 

 

 

SOLUBILITY (moles of Ne/cm3 of melt})
e N
o o

/

 

 

 

 

 

 

.0 .- 080 100 150 200 - 250  3.00
SATURATING NEON PRESSURE (atm)

o

3.0

  
 

 

 

20 Fig. ?.3.13. Solul:llify of Neon in Molten NoF-KF-

 

 

 

 

 

 

 

 

"-"_LIF (11.5-42-45.5 Mole %
o Solublllty of HF in Molten Fluorldes
o | : s _ , " J.H. Shaffer |
O - 1050 CThoo o 200 260 Meosurements are being made of the solubility

o .SATURATING HELIUM PRESSURE (atm) . o of HF “in various. molten fluoride solvents as a

o s TR ald funchon of temperature, pressure, and solvenf
. Fig. 23.12, Solubilny of Hellum in Molten NuF-KF- composition. The effects of composition of mixtures
LiF (11,5-42-46.5 Mole %). in the NaF-ZrF4 system are shown graphically in

93

 
 

 

 

 

-~ MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED

ORNL—LR—DWG 27923

TEMPERATURE (°C)
_; 800 700 600
3X10 —

2 ey ny

= 8000 cal/mole

.N€ON

8900 cal/mole

K [moles of gas/(cm® of melt) (atm)]
L]

o

 

9.00 40.00 11.00

10Y/7 (°k)

12.00

Fig. .2.3. 14, Temperature Dependence of Sclubilities
in NaF<KF-LiF (11,542-46,5

of Helium and Neon

Mole %).

Fig. 2.3.15. There is a remarkable increase in
HF solubility with increasing NaF concentration.
Preliminary results obtained for the solubility
of HF in NaF-KF-LiF (11.5-42-46.5 mole %) are
presented in Fig. 2.3.16. Since the liquid structure
of this solvent might be assumed to approximate
that of molten NaF, the value of the solubility of
HF in the NaF-KF-LiF system may be used at
0 mole % ZrF, in Fig. 2.3.17. This curve was
.constructed from solubility measurements of
solvents containing 19.5, 35, 40, 47, and 55 mole %
- ZrF ;. The regions from 0 to 19.5 and from 19.5 to
- .35 mole % ZrF, were interpolated because of the
- high melting points of the corresponding NaF-ZrF
mixtures. The heat of solution of HF as a function
- of solvent composition is shown in Fig. 2.3.18.
_ The curves presented in Fig. 2.3.17 reveal a
50-fold increase in HF solubility as the,Zr'F“
concentration decreases from 55 to 0 mole %. The
heat of solution increases about threefold in the
corresponding composition range. Similar corre-
lations will be attempted as other solvents are
sfudled
A study of the solubility of HF in LiF- BeF2

mixtures was mmated and the resuhs obtained to

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-5 UNCLASSIFIED
(x10°%) . ORNL-LR-DWG 27924
4.5

.’.
4.0
35,

I"'"!.

E

2 30

75-‘ 800°C

s

")

§ 2s

~

w . )

T 1

B \600°C :

o
3 20
£ .

—Jd

X
1.5 \ "o .

1.0 AT N\
\. \
.. ™
'\\.\
N\
0.5
10 20 30 40 50 60

ZrF, (mole %)

Fig. 23.15. Henry's Law Constants for HF So!filfility
in NoF-ZrF‘ Mixtures as a Function of Solvent Come
position.

date for the LIF-BeF (51-49 mole %) mixture
are shown in Fig. 2.3. 39. The heat of solution of
HF in LiF-BeF, (51-49 mole %) as calculated from
the data presented in Fige 2.3.19 is —4100 cal/mole.
Both the HF solubility and heat of solution values
for the LiF-BeF, (51-49 mole %) mixture are of the
same order of magmtude as the values: obtained
with corresponding NaF-ZrF, mixtures. Additiona!
measurements of HF solubility in mixtures in the
LiF-BeF, system are being made. '

Solubilities of Flssmn-Product Fluondes
) W. To Ward
lnveshgahons are under way for determlnmg the

solubllmes, of some fission-product fluorides in
various BeF ,-containing solvents. The radicactive

 
o

 

 

 

U

. -..7 UNGLASSIFIED

P
1 Y
(x10 ) ORNL-LR-DWG 27925

8.0
7.0 //
6.0 ,

1/

17

2.0 7

 

 

 

 

 

 

X [moies of HF /(em? of mert) (atm)]

 

 

 

 

 

 

 

 

9.0 9.5 10.0 10.5 o s
10%7 (°K)

Fig. ‘2.3.16. Henry's Law Constants for HF Solubility
in NaFeKF-LiF (11,5-42-46.5 Mole %) as a Function of
Temperature. '

UNCLASSIFIED

 

 

 

 

 

 

 

 

 

 

 

 

ORNL—LR—DWG 27927
15,000 1 I I I i
o-AH FOR NaF—-KF—LiF (14.5-42.0—46.5
. mole %) . _
S 0-%;
'~ 10,000 [———
o
o ; .
E
©
x
Q-
- I 5000
-0 40 .7 20 730 0 .40 50: .60

© ZrF, (mole %)

Fig. 2.3.18. Heat of Solution of HF in NaF-ZrF,

Mixtures.

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED

— 5
(x40 ) ORNL-LR—DWG 27926

100

LIQUID PHASE
====S0OLID PHASE

50.0

040 FROM MEASUREMENTS ON
NoF-KF-LiF (11.5-42-46.5
20.0 mole %)

10.0

5.0

K [motes of HF /(cm3 of meit) (atm)]

20

05

 

03 ‘
0 {0 20 30 40 50 60
Zrfy {mole %)

Fig. 23.17. Henry’s Law Constants for HF Solubility
in NuF-ZI'F4 Mixtures,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-6 UNCLASSIFIED
(:(;0 ®) ORNL—LR— DWG 27928
9 V //
8 — L e
e
S
w9
:'t ’
s
gt
9 95 1 400 405 - 40 1.5
0% ek

Fig. 2.3.19. Henry's Law Constants for HF Solubility

in LiF-BeF, (51-49 Mole %).

95

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

tracer techmques described in a prewous report9 .

are being used.

CeF in NaF-BeF,. — The values of the solu-
in two_ NaF- BeF2 mixtures were

bility of CeF
determined at ?our points in the temperature range
420 to 720°C. The data are summarized in
Table 2.3.8. '

YF in NaF<BeF,. — Values obtained for the
-solublllty of YF_in NaF-BeF (61-39 mole %) are
presented in Table 2.3.9. The fotal Y+**.in the
‘system, as calculoted from the weights of YF
and solvent added at the beginning of the experi-
‘ment, was 6.0 wt %. No satisfactory explanation
can be offered for the amount apparently found in

 

"W, T. Ward et al., Solubility Relations Among Some

Fission Product Fluorides in NaF-ZrF -UF (50-46-4

Mole %), ORNL-2421 (Jan. 15, 1958). 4

Table 2.3.8. Solubility of CeF; in NaF-BeF,
o Solvents

 

Filtrate Analysis

Ce Cer
(wt %) {mole %)

 

Temperoture

(°C)

 

Solvent: 81 Mole % NaF -39 Mele % Ber*

718 2.58 0.83

639 1.29 0.41
517 0.39 0.12

424 0.15 0.047

Solvent: 56 Mole % NaF—-44 Mole % PeF,*

716 2.86 0.93
627 1.38 0.44
520 0056 018
424 026 0,083

 

*Composition of solvent based on chemical analysis.

Table 2.3.9. Solubility of YF, in NaF-BeF,
| (61-39 Mole %)

 

 

 

Terhpérature Filtrate Analysis
e Y (w %) YF, (mole %)
e 7.9 4.3

617 7.6 4.1
. s2 3.8 1.9
4 22 - 1.1

 

96

“LiF-BeF,

the first two filtrates bemg higher than the calcu-
lated amount.

CeF3 in. LiF-BeF,. - Expenments for de-
termining how a varlcmon of the lithium-to-beryllium
ratio in LiF-BeF. solvents affects the solubility
of CeF, are in progress. . The results obtained
thus far are given in Table 2.3.10 and shown
graphically in Fig. 2.3.20. The solubility of
CeF, appears to be ot a minimum in the mixture

_contalnmg about 63 mole % LiF. The saturating

phase in all cases was found, by mucroscopy, to
be pure CeF ..

CeF. and Lch in LlF-BeF » — The solubllmes
of Celg and LaF, in the presence of each other in
(52.5-47.5 mole %) were determined
with the use of two radicactive tracers. The total
composition of the system was calculated to be
2.05 mole % CeF , 4.41 mole % LaF3, and 93.54
mole % solveni. The results are shown in
Fig. 2.3.21, in which the logarithm of the solu-
bilities (mole %) is plotted against the reciprocal
of the temperature.

UNCLASSIFIED
- ORNL—LR—DWG 27929

 

 

 

 

 

 

 

Cefy IN FILTRATE {mole % )}

 

0.4

 

 

0.2

 

 

 

 

 

 

 

 

 

30 40 50 60 - 70 - BO 90
' LiF IN SOLVENT (mole %} .

Solvents.

Fig. 2.3.20. Solubility of CeFa‘ in Vfirious LiF-BeF2

 
 

Q\

Al

 

PERIOD ENDING JANUARY 31, 1958

Table 2.3.10.-_‘.‘S§|ubi|ity of CQF3 in Various ,LI_'F&Ber Solvents

 

Solvent Composition

 

Filtrate Analysis

 

 

Filtrate
(mole %) Temperature Ce CeF3
LiF BeF, (°C) (wt %) (mole %)
51.6 4844 726 6e47 1.80
615 3.12. 0.83
501 1.09 0.284
407 0441 0.105
53.8 4642 N7 6441 1.76
632 2.98 0.79
511 1.04 00269
412 0.367 0.094
5647 43.3 723 6409 1464
612 2,42 0622
508 0.86 0.219
425 0.328 0.082
5943 40,7 704 4496 1429
627 2.66 0:677
544 1.14 0.286
476 0,515 04128
62.7 37.3 720 5.77 1.50
: 643 2.91 0.73
556 1426 0.309
467 0.442 0.108
6640 34.0 729 6496 1.79
656 3.72 0.92
556 1.34 0,322

476

0.535 0.127

 

" Order of Oxlde Precipitahon in Fluoude

The precnpltation of oxudes of flssmn products

Salt Melts
M Blander

" The rule for the relative preélpltctlbh'of oxides

"-*from a melt is also ihe rule of stability. The order -

of precipitation of Ce, 203 and BeO in the LtF-BeF

. from fluoride melts is bemg sfudled -as one of the -
- possnble methods for the purlflcuhon of moltensalt =
~Although the. detalled
. theoretical anulySIs of such™ precipitation s
relatively complex, snmple thermodynumlc con-:"{";}
~ siderations ¢an be opphed in order to calculate -
the relative order of precipitation of oxides, . As;fi_,

on illustration of the principles involved, calcu-

reactor “fuel - mlxiures.,

lations are described here for the oxide precipitation
of U4t Ce**t, and Be** from an LiF"-BeF2
mixture containing UF , and CeF .

melt, for example, is mveshgated by comparing the'--_ ,
1hermodynam|c stability of the pair of compounds

on the right of the following equation with that of

“the palr on the left s:de of the equaflon'
(‘) Ce O + 3BeF2@2CeF + 3Be0

‘:At temperotures at which the oxlde is.a sohd and
- precipitates in pure form, the oxide that precnpltotes,

first should be a member of the stable pair. The
procedure is, then, to calculate the free-energy
change, AFS, for the pure solids of the reaction
represented by Eq. 1. The ratio of the activities

97

 
 

 

 

 

"('2) “AF® = —RTIn K =

= will precipitate.
' ~140 kcal was obtained by using the values of the

RARE = EARTH- FLUORIDE IN FILTRATE {mole %}

UNCLASSIFIED
ORNL-LR-DWG 27930 .

N TEMPERATURE (°C) e
700 600 - 500 - 400 -

TOTAL COMPOSITION OF SYSTEM:
+2.05 mole % CeF3
© 4.41 mole % LoFs
-93.54 mole % SOLVENT

S
04’0»

05

- 0.2

 

- 04
0.05
0.02
0.01
10 it 12 13 19 15
104/ 7 (°K} ’

Fig. 2.3.21. Soiubiliiies of (;eF3 und'Lan'When Both
Are Present in LiF-BeF2 (52.5-47.5 Mole %) Solvent,

"(BOSed on the solid as the standard state) of the

two_ fluorides involved in the reaction when the
solvent is in equilibrium with the pure solids
Ce,0; and BeO smultaneously can then be calcu-

lated iy usmg Eq. 2:

‘. a2
: - CeF3
~2.303RT log ——— .
. 613 : .
Ber-

ifl'the equilibrium ratio a%;[-_._ /“g'eF is greater

- than the ratic in solution, BeO will precipitate,

and, if it is less than the ratio in solution, Ce 0,
‘A value of AF° for Eq. 1 o

'9.8

@ e

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

 free energles of formahon given in Table 2.3.11,
~and from Eq. 2 it was fcund that

N 140, 000/4570
CeF‘s/aBer ]0 ’

= 1029'8 = 63 x 1029 .

Table 2.3.11.- Free Energtes of Formufion-of Fluorides

" and Oxides of Be'", Ce***, and U4* ar 1000°K*

 

Free Energy of Formation -

 

: (kcal/_niole)
~ Fluorides -
'B;-szi_' | ~196
_‘Cei'-"3 . ' .'fi ., -—.360
UF4 - . ‘ | -373
| Oxides _ _

BeO 123
(:e203 _ B —-361
‘U02 -218

 

*Data reported by A. Glassner, A Survey of the Free
Energies of Formation of the Fluorides, Chlorides, and
Oxides of the Elements to 2500°K, ANL 5107 (Aug. 1953).

The activity, based on the soiid as standard
state, of CeF, in a 1 mole % solution in LiF-
BeF, would be of the order of 10-3; the activity,
based on the solid as standard state, of BeF
in a 70-30 mole % mixture of LiF- BeF, would be
of the order of 10-2; and the’ eshmated ratio of

CoF /aBer is about 1. Since the calculated

eqmllirlum ratio of activities is much greater than
the estimated ratio, it may be concluded that BeO

~ will precipitate before Ce ,0, precipitates. Since

the calculated equrllbrlum I'CI‘I‘IO is so large, most

~of the BeF “will probably precipitate before any
-Ce20 precupltotes,

These conclusions are true
only af the BeO and Ce 03 form no compounds and
do not form solid solutions. if Ce, 05 precipitates
readily it must be as a compound or soiid solution
with BeO.

“The order of precnpn‘ahon of U“+ and Ce"""+ may
be studied in a similar manner, The AF°_for the
reaction ' ’

-4 3UF + 2Ce203<———-3U0 + 4CeF

is —253 keal,

‘and the equ:ilbrlum ratio of the

-

 
 

 

 

an

s )
(5) “cer/“UF4

activities of CeF3 and UF4 calculated from this
quontlty is, at 1000°K,

]0253,000/4570

= 10553 = 2 x 10%5 .

The activities of UF, and CeF3 are of the order

of 10=3 for a 1 moTe % solution of these two

substances, and a crude approximation of the ratio
aé.F /a} ¢ would give a value of 10-3. Since

this is much smaller than the ratio calculated in

Eq. 5, it is concluded that UO, will precipitate
almost completely before any Ce,0, precipitates.
This conclusion is subject to the condition that
only pure Ce,0, or UO, can precipitate. These
sample colcufahons |I|ustrate the utility of even
crude approximations to the activities of com-
ponents of salt mixtures. The same conclusions
would have been obtained even if the individual
activities had been in error by an order of
magnitude,

Chemical Reactions of Oxides with Fluorides

in LiF-KF
J. H.'Shaffer

Selective precipitation of oxides may be useful
in the development of o svitable scheme for the
reprocessing of molten salt reactor fuels. Ac-
cordingly, the solubility and some selective
precipitation reactions of a variety of solutes are
being studied. In order to avoid complications,
the simple binary mixture of LiF and KF was
chosen as a solvent. _ :

PERIOD ENDING JANUARY 31, 1958

On the basis of direct and indirect information
gained from the experiments completed thus far,
the following useful generalizations can be made
regarding the solubility of oxides of uranium,
zirconium, hafnium, rare earths, alkaline earths,
and alkali metals in this solvent:

1. Uranium, zirconium, and hafnium precipitate
as the dioxides and are the least soluble.

2. Rare earths precipitate as R203 and have
very low solubilities.

3. Beryllium and magnesium oxides are slightly
more soluble than the rare earth oxides but are
still quite insoluble.

4. Barium, strontium, calcium, potassium,
sodium, and lithium oxides are more soluble than
the oxides specified in items 1, 2, and 3.

Quantitative measurements of the solubilities of
the relatively insoluble oxides are difficult to
make, because the solubilities are so low as to be
below the direct limit of detection of the radio-
chemical method used. However, extrapolations
have been made in order to estimate the concens-
tration of the particular metal ion remaining in
solution in the presence of some 25% excess
precipitating agent (usually CaO or Na,0,). The
results are presented in Table 2.3.12.

Several separations of solutes appear to be
possible on the basis of the selective precipitation
of oxides obtained in the experiments. The
results of the experiments are presented in
Figs. 2.3.22, 2.3.23, and 2.3.24, which show the
concentration of solutes that remained in the
liquid as the system was titrated by the stepwise
addmon of CaO. For example, Figs. 2.3.22 and

Tuble 2.3.]2. Concentrntlon of Meful Ion in Molten LIF-KF at 600°C in the Presence
' of Excess Pteclpltuting Agent :

 

: Metul Found -

 

- Méfdl‘,,'lrénr -» p.e'cipifa-,;,.g- phos;{ o pepiiA | 'A;.;:!_yoicol Method Used
7 i‘;f;;,,_(u,rn“ S r_rz.no2 -,éb.OOOBV o : . Radiochemical -
Héfnium L S Hf02 . <0.0008 = P S Rudiqchemicul
- _'ur'anm.fif S uo, . <0.0085 . . Chemical
: ‘CeriUm CEo .,Ce203 <0-0'I STl 7 - Rodiochemical
Beryllnum,'r“_"’ L ; BeO <0.02 . Chemical
- Barivm - . Soiuble to extent cdded >6.5 R fi . Radiochemical

 

*Compositions of saturating phases calculated from indicated end points of titration curves obtained with CaO as

the titrating agents

99

 
 

 

 

' MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

2,3.23 show that uranium is precipitated essentially

quantitatively as UO, before any of the Ce203 is
formed.
experiments wete determined radiochemically,
while the uranium concentrations ‘were determined
by chenical analysis. In the titration of the

 

mixture ZrF CeF, with CaO, as shown in
UNCLASSIFIED
5 ' ORNL-LR -DWG 27931
e
4 ._———.-"— R

 

_Ce (DETERMINED RADIO- _|
©7 CHEMICALLY)

 

N\

\

0 0.1 0.2 0.3 04 0.5 0.6
' EQUIVALENTS OF CoO ADDED )

 

\

' METAL FOUND (wt %)
UF, END POINT——- — .|

U (DETERMINED
"CHEMICALLY)

\
/

 

 

 

 

 

 

 

 

 

 

 

Fig. 2.3.22. Titration of CeF3 and UF4 with Ca0 in
LiF-KF (50-50 Mole %).

UNCLASSIFIED
ORNL-LR-DWG 27932

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5
| | l
o —U(DETERMINED CHEMICALLY)
4 \\ UF, END POINT ———=
@
z3 ‘ N
o ) ®
=
=2
O
L.
32
'.—
b N
I
1 |————— Ce {DETERMINED
. / RADIOCHEMICALLY)
¢ _ T Q T '<\ :
. @ . i
0 , . .|
0 1005 O0t0 - 0Of5 0.20 0.25

" EQUIVALENTS OF CaQ ADDED.

Fig. 2.3.23. Titration of CeF, and UF, with CaO in
LiF-KF (50-50 Mole %).

100

The concenfratmns of cerium in these -

| Flg. 2.3.24, the seporatlon does not appeor to be

quite as sharp as in the case of the UF ,-CeF
mixture,  as shown in Figs. 2.3.22° and 2.3.23.
The htrahon with CaO of ZrF in the presence

of nonradsouchve CeF3 is shown in Fig. .3.25

UNCLASSIFIED
ORNL-LR-DWG 27933

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0 —
9
Zr (DETERMINED CHEMICALLY) .
1.5 : ,
3 \
= L
E‘ I er4 END POINT —=
2 10 _
4
=
W
E.
Ce{DETERMINED
RADIOCHEMICALLY)
0 - 0.05 - 040 0.45 0.20

EQUIVALENTS OF CaO ADDED -

Fig. 2.3.24. Titration of Cer
in LiF-KF (50-50 Mole %).

and ZrF4 with Ca0

UNCLASSIFIED
ORNL-LR—DWG 27934

 

2'06 I :

  
 

aa—2Zr(DETERMINED RADIOCHEMICALLY)

 

 

 

 

5 , ZrF4 END POINT —=
_ \ -
B2 | ‘\
:
2
> e
S 10 ‘
=
2
=
Q
£
xz .
N 05 =\

 

 

 

 

 

 

 

o 005 ‘040 015 0.20

‘EQUIVALENTS OF CaO ADDED -

Fig. 2.3.25, Titration of ZrF

in the Presence of

CeF with Ca0 in LiF-KF (50 50 Mole %)

 
 

 

 

7

In order to determine the concentration of zir-
conium radiochemically, the zirconium was labeled
with Hf181 and the hofnium was Gssumed to
behave like zirconium. The titration curve shows
the assumption to be sound.

The titration of the mixture Cer-BeF with CaO
is shown in Fig. 2.3.26. From the curve it may be
seen that cerium and beryllium are precipitated
simultaneously but not in definite proportions;
they probably precipitate as o solid solution.
The results are summarized in Table 2.3.13. [t
has been suggested!? that the results of this
experiment indicate that an attempt should be
made to precipitate cerium from the solution by
adsorption on an excess of BeO. Accordingly,
attempts are being made to precipitate not only
cerium but also uranium.

Lithium Recovery from NaF<KF«LiF Melts
R. A. Strehlow

The possibility of recovering the economically
valuable materials from molten salt reactor fuels
is of concern in the selection of a particular set
of fuel components. The mixtures of alkali.metal
fluorides which have been considered suffer to

 

METAL FOUND (meq/ 100¢ of solution)

= -l o

10g, 3, Sturm, private communication to J. H. Shaffer.

 

 

 

 

 

 

 

UNCLASSIFIED
350 ' . ORNL—-LR—DWG 27935
1T 17 T 7]
) & CERIUM REMAINING IN SOLUTION
300 | — - {DETERMINED RADIOCHEMICALLY) : )
: 'O BERYLLIUM REMAINING IN SOLUTION . - -

. (DETERMINED CHEMICALLY) - '
250"

S R "0 °TOTAL OFCERIUMVAND BERYLLIUM

b T e msowrron S
200 f——f-— \ e
150  }——feme ‘:_\\\ — 1
100 \

 

 

obl 1 L1 T

 

 

 

 

 

 

 

 

 

"0 . 50 100 450 - 200 250 - 300 350

CaO ADDED (meq/100g of solution)

Fig. 2.3.26. Titration of Cer and Ber with CaO
in LIF-KF (50-50 Mole %).

PERIOD ENDING JANUARY 31, 1958

Table 2.3.13. . Titration of CeF, ond BeF, with Ca0
in L‘I.F-KF (50-50 Mole %)

 

Metal lon Found in Liquid
(meq/100 g of solution) CaO Added

(meq/100 g of solution)

 

 

Ce ++4 Be ++

112 197 0
9 186 21.0
80 184 39.8
' 58 182 6142
40 175 82.0
27 7 102.8
21 118 164.0
14 42 225.9
0.5 " 287.0
006 4 342.0

 

some extent from the restriction that isotopically
pure Li7 must be used. The discarding of the Li7
contained in these fuels after the maximum
allowable poison concentration had accumulated
would place a moderately serious cost burden on
the power produced from a lithium-containing fuel.

Either of two approaches may be used to increase
the utility of the fuel. Fuels from which the

lanthanides can be removed can, of course, be

reprocessed and reused. This presumably is the
case for LiF-BeF y~bose fuels in which the

~ principal residual (nonlonthamde) poisons would
. be primarily elements which are more noble than
~beryllium. - if, on the other hand, the rare earths

cannot be conveniently removed, the lithium must,

"'_'perforce, be removed from the balance of the fuel.
- This is the case for alkali- fluoride melts con-
. taining comblncmons of Inhlum, sodwm, und

potosswm fluorides.
1t would ‘be desirable to be able to remove the

lithium from NoF-KF-LiF . mixtures by o simple

hlgh-tempemture technique and for the product to
be in a_concentrated form ond free from rore-earth
contamination. It may be seen from Table 2.3.14,

 which "shows the free energnes‘of.formahon at
1000°K _of some halides, that somarium halides
“are - considerably more stable than - the corre-

sponding alkali metal haolides. Although, among
the alkali metal halides, lithium forms the most
stable fluoride, potassium forms the most stable

101

 
 

 

 

* Table 2.3.14. Free E‘nergie'srof Formation of Halides* "

 

~AF ot 1000°K (keal /mole) -

 

 

W | )
B F= SCl= BT T
Cut o e 802 700 564
et 1126 767 683 560
K 08 816 752 6.5
sttt 147 657 547 393
Csett 1250 860 730 '61’.0_-':

et '97.5 . 42;3 299 '12.7 |

,Be

 

 2 *L- Brewer et als, N’atl. Nuclear Energy Ser. Duv.IV ‘

198 (1950). .

: _'cllwlorlide,. -.brb‘mi'de, ,ond. iodide;" t Seems._—fo be

possible that the reaction described by

(6 LiF + KS==KF + Li

 might be made to occur if, for example, potassium
chloride - or bromide were added in sufficient

amount to a mixture of the fluorides. Although
sodium would be extracted to a greater extent than
lithium, its presence in small amount is not a

‘serious consideration. The equilibrium constant

for the reaction (Eq. 6) can be calculated, since

4 19K F

(7)) AF°=-RTh K, = -RT In— ,

Akq 1

- where the a's represent the thermodynamic ac-
~ tivities of the substances symbolized by the sub-

scripts, R is the gas constant (1.98 cal/moleK),
T is .the absolute temperature, and AF° is the

- change in free energy- per mole when the process
- is carried out with-all substances maintained at
| umf ochvnfy. Thus at 1000°K,

]5 700 cal/mole ( ~1.98 cai/mole.°l() % -

| x 1000°K X In Ky o

UK, =36 %1074 .

| ‘_The ‘principal _drifficulfy' in- carrying out calcu-

lations for more complicated cases (for example,

02

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

mixed fluoride-chloride systems) is in-the esti-

_‘.'mahon of the values of « /aK «and aKF/“LlF
; ‘_for the ‘metal phase and salt phuse, respectively,.
" as functions of composition. The ratio, a, Jay,

may be ‘assumed to be equai to N, /N, where

.NL, is the mole fraction of ||fh|um in the metal

phase.  This is equivalent to assummg |dea|

~behavior. in the metal.

“In the salt phase the ratio of activities may be

- -colculafed for. KF-LiF-KCl as a function of the
- amount of KCl added. The methods which have
been used for the calculation of activities in

‘multicomponent . systems  are described below.

- “Use is made of the approximation.

NN
My X7

1% N - i L2

T . +

( Mt ,2x7)'

 

where N, and N
oM 1
M ’ cnd X ", and the expreésmn NEM+'+ NEX_,
represents the totol number of moles of all positive
and negative ions. This approximation'! may be
expected to serve only at extremely high temoer-
atures, since, clearly, no distinction is drawn
between positive and negative ions. The calcu-
lation lmplles a gas-like character of the melted

salt. Temkm used the relation:

 

 

(9) ‘.IIM X, = ;
1“1 |
| Tswt Zx*

where 7 "‘+'--dhd n_
.
of the posmve “and negatlve ions, respectlvely,

ot is the number of moles of all the positive
M

ions, ond-n -_is the number of moles qf-qll the

 negative ‘ions. - Here the  "activity’”’ of the in-
dividual positive ion is related to the other

pOSlflVe ions instead of fhe ions of bofh charges.’

 

- _(1938)
]2M Temkm, Acta Pbyszcocb:m. URS.S'. 20 411

P. Harasymenko, ' Trans..» Famday .' .S‘oc.: '34 ,1245

Q 945)

_ are the numbers of moles of .

are the nurnbers ‘of moles

L

 
 

 

 

 

"

The Temkin definition of mole frfiéfibn provided
the point of departure for Flood, Forland, and
Gr;otheim,' who, from theoretical consnderohons,
constructed a method of caleulating the activity
coefficient. Their relation may be symbolized by:

’ mn
(10 Iny,, =~ z L N N AFS/RT

where y, . is the activity coefficient in a molti-
component system containing m cations and n
anions, and AFS. is the standard free energy of
formation of component M _X. by the simple meto-
thetical reaction - which also produces M, X,.
This snmphfles for a sample ternury recnpfocal
system fo
Iny,; = -Ny N AF‘;‘ZXZ/RT .

The octivity is calculable, therefore, by the
equation

-(N, N, AFS /RT)
Mg X, " MaX,y

For the system under consideration, the meto-
thetical reaction that produces the various com-
ponents is

LiF + KCI'-——?*KF + LiCl

which has an assoc:ated AF° at 1000°K of

17 1 kcul/mole. - '

_ The ternary mlxture LlF-KF-KCl is dlsployed;
-_l'schemutlcally as o reciprocal system in Fig, .
2.3:27. Point 1 in Fig. 2.3.27 represents a mixture -
. that is 50 mole % LiF and 50 mole % KF. This -~
,_,-';composmon is “near the eutectlc composnhon for[ '5-**'
<+ . the KF-LlF binary, which contains 52 mole % LiF, .7t
~ and was" chosen as ‘a representative fluid wuth, LT
properhes ‘similar to - those of- _potential  fuel
“mixtures, - The' line ‘AB is the locus of compo- -
sitions formed upon addition of KCI to the’ original -
"composihon. This dine is dmded into four equal .
- parts by pomts 2,3, and 4. The calculation of - .
- ay/ag from Eq. 7 is" “shown below for ‘each of
"these compositions at ‘1000°K. ~The pertinent

 

13, Flood, T; Forland, and K. Grjotheim, Z. anorg.
u. allgem. Chem. 276, 289 (1954).

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED
~ ORNL—LR—DWG 27936

LiF ' A . KF

 

 

 

 

 

 

LiF
LiCL
: B
LicL KCL

Flg. 2.3.27. The Ternary Mixture LiF-KF-.KCl Dis-
played Schematically as a Reciprocal System.

expressions are:

=N _N N N AF,~/RT
“LiF TN e exP( k* ci- ke )'

agp = N o N__ exp (NL#NcI_ AFyc/RT) -
For point 1:
a_ig = agp = 0.5 .
For pomt 2:

“LiF = 0.375 X 075 X .
C xexp [0.625 x 0.25 x (+l7.l)/l.98] ,
. 0.625 X075 x

o exp [0.375 x- 0.25 x ( 17 l)/l.98]

“Kr- . 0.625

==

AL T 0375
B = 0.2] . .'

 

 

exp ('-"-'wl'.35 — 0.81) o

For pomt 3:
@ F = 0.25 x 0.5 X
x exp 10,75 x 0.5 x (+17.1)/1.981 ,

103

 
 

 

 

MOLTEN.SALT REACTOR PROGRAK PROGRESS REPORT

. aKF = 0.75 X 0.5 x
X exp [0 25 x 0.5 x (=~ =17.1)1 98]

“_KF '

 

‘ =3 exp (~3.24 - ];08)
ik

| - 0.108 .
For point 4: |
e = 0,125 x 025 x
- xexp [0.875 x 0. 75 x (+17.1/1.93)]
agp = 0.875 x 0:25 %

X exp [0.125 x 0.75 x (~17.1/1.98)1 ,

KF

i

 

- : 7 7 exp' ("'5-67 - 008])
C%LiF o

7 exp  (-—6.48)
0.0106 .

¥

For point 5:
Ne, =0.875 ,
a f = 0.0625 x 0.125 x

X exp [0.9375 x 0.875 x (+17.1/1.98)],

aK_F -0o9375 x 0. ]25 X
X exp [0.0625 X 0-875 X ("1701/1 -98)] ’

- aKF

 

| = 15 exp (~7.08 — 0.47)
“LiF L L

= 15 exp (—-7-5_5)'

= 0.0078 .

Since K for the reaction (Eq. 6) was known to be
3.6 x 10=4, values for the ratio ay ;/ay could be
calculated. The calculated values are shown in

Table 2,3 15.

 The formal significance of these calculations
which ‘involve the Flood, Forland, ond Grjotheim
relation may be clarified by consuierlng the adding
of KCI to LiF until a particular composition in

‘the ternary diagram of Fig. 2.3.27 is reached,

namely, the very center of the diagram, where

N <N . =N =N =0 ,
+ K* F~ C!_OS

and therefore

The equilibrium constant for the reéction’
KCl + LiF = LiCl + KF
may‘ be expressed as

NerNiict  YkFYLic
K = KNK .

a ')/ ’
NKCINLuF YKCIYLIF

 

Tub.le 2.3.15. The Ratio a, /d at 1000°K for the Reaction K + LiF ;KF + Ll ‘upon Addltlon
: of KX to KF.LiF (50—50 Mole %) : :

 

Ruflo aLi/él'( for

 

 

 

KX - o ‘ _ : , '
‘ (mo|e %) X = CI= X = Br™ i - X =1
S ArF° = 17.1) (AF® = 20.8) - AF° =225
" 0.0 | | 36 % 1074 36 x 1074 36 x 104 -
025 19 x 1073 2.7 x 10~3 3.7 x 1073
0,50 3.4 x 1073 22 x 1072 3.5 x 102
075 3.4 x 10~2 14 x 10-7 2.6 x 10~}
o875 46 x 1072 2.2 x 1071 49 x 107

 

104

iy

 
 

 

o "NL /N

 

where K_ is determined from the expression

AF°® = —RT In K .

Since Ky is unity,

A value for Ky of 6.2 x 10~% at 1500°K is then

obtained from the expression

~17,100 = RT In K, .

As a crude approximation it may be assumed
that LiF and KCI (the stable pair) have equal
activities. This permits an estimate to be made
of the activities:

1 1
YKF = YLicl =—— S———
¢ YKCl  YLiF
]
vk = (62 x 105)/4 = 0.089 = —— ,

YLiF

acl = aLiF = 2.8 4

et = ekp = 0.02 .

It should be noted here that for this example a

measure of the error involved in the calculation

may be seen, since an activity in excess of unity
(using the solid as the standard state) would
indicate precipitation or at least an immiscibility
gap. Since neither of these is known to occur at

1500°K, it would be expected that these valves
- may be in error by, perhaps, factors of 3. This “previously, 15

- would result in an_error factor of 10 for fhe ratio ’

aLi/aK- S

rcmos. .

equnpped with a “nickel filter. -

The injector was then pressurized with argon to
force the molten metal through the filter into the
reactor, where the potassium was equilibrated at

PERIOD ENDING JANUARY 31, 1958

450°C with the LiF-KF mixture. The mixture was
then cooled to 150°C to permit the salt to solidify,
and two samples of the molten potassium were
taken.  Although the sampling temperature was
below the consolute temperature for the Li-K
system, it was suitable for any small ratios of
ay /axs A modified metal-sampling apparatus 14
was used which permitted samples of liquid metal
to be obtained under an inert atmosphere. Two
quantities of KCl were then added in amounts of
100 and 300 g. The mixture was equilibrated, and
the metal phase was sampled after each addition in
the manner described above.

Analyses gave values for the three determinations
of Ny ;/Ng = 10=% and of N__ =0, 0.16, and

0.35, respectively. One equilibration of a mixture

of KI-Lil-K containing less than 0.05 mole fraction
‘of fluoride was carried out, and a value of
| NLl/NK

0.35 was obtained. These results
indicate that the estimates described here and
presented in Table 2.3.15 may be accepted as
being of the correct order of magnitude. This
leads to the conclusion that, at probable Li7
price levels, this process is not economically
feasible.

CHEMISTRY OF THE CORROSION PROCESS

G. M. Watson F. F. Blankenship
G. J. Nessle

Aéiivity Coefficients of CrF, in NaF.ZrF,
C. M. Blood
The activity coefficients of.CrF, dissolved in a
molten mixture of NaF-Z¢F 4 (53-47 mole %) are
being determined by using techniques described

The - activity coefficients of the
metal fluoride solute are determined from experi-

-~ mentatly ‘measured values of equullbrlum quohents
o of the reachon' - ' :
. , Experlmental work has been currled out for the o
- purpose of - ascertaining the “magnitude of the
L Samples consisting of 300 g of - .
- L:F-KF (50-50 “mole %) were equilibrated in a.
“nickel reactor with :about 70 g of potassium. The ich
potassium metal addition was prepared by melting
it and casting it into o nickel injection device
The injector was
attached to the reactor, ‘and, the - potassium was
‘remelted by application of ‘a flame to the injector.

et

o

MF,(d) + Hz(g)v—i‘ M(S) + 2HF(g)

The results of exper:mental measurements on

ut 850°C are summarlzed in Table 2.3.16.

 

”G coldberg, A. 5. Meyer, Jr., and J. C. White, The
Samplmg of Alkali Metal Systems with tbe Modzfzed
MSA Sampler, ORNL-2147 (Sepf. 10, 1956) .

l5(:. M. Blood “W. R. Grimes, oand G. M.’ Wafson,
Activity Coeffzczents of Ferrous Fluoride and of Nickel
Fluoride in Molten Sodium Fluoride~Zirconium Fluoride
Solutions, paper No. 75, Division of Physical and
Inorganic Chemistry, 132nd Meeting of the American
Chemical Society, New York, Sept. 8-13, 1957.

105

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

Table 2.3.16. Eq'uilibrium. Quotients at 850°C of the Reaction CrF,(d) + H,(e) = c'(-_’). + 2HF(g)

 

 

Pae Xcrer, K Pur Xcrr, K%
(atm) (mole fraction) x (atm) (mole fraction) x
x 10=3 x 10~2 x 10~3 x 10=3 - x 1072 x 10™3
9.55 - 9.48 0.97 6449 3405 , 1.39
9463 942 0.99 6463 3.1 ' 1042
1078 = 9.27 1.27 643 3.19 131
" 10,54 - 9.14 123 7.98 ~ 3.87 C 66
10455 . 9.08 1424 816 3.91 172
11.19 9442 134 673 3.82 | 1.19
.01 9.25 1:32 6491 396 L2
11.26 9.29 1.38 7:30 | 4.04 133
11.85 9.42 1.51 7475 ' 409 - 148 -
1178 935 1.50 740 4.07 . 1435
11,64 937 . 1446 7.83 4407 " 1.52 -
1180 973 1,47 845 425 170
12,04 986 1449 7:96 423 1.51
12.51 | 9.84 1.61 8.58 427 1.74
12,57 9499 1,60 3.88 1.26 | 1.20
11,65 . 973 1.41 3.78 1,14 1.26
12.06 952 1.55 3.91. 1.03 1450
12,58 9.50 1,69 4,33 1.18 1.59
13.51 9:86 1.88 4.83 : 1.43 1.64
13.39 3 9.90 - ~ 1.84 536 1.53 1.89
5.12 2.63 1.00 4,81 ST 1448
6443 3.04 1.37 4.63 1.63 1432
699 3.04 1462 473 1.63  1.38

 

e

Av 145 £ 0.17

 

'*Kx = PEIF/XCr'FzPHz, where X is mole fraction and P is pressure in atmospheres.

This tabulation shows the experimentally de-
termined partial pressure of HF and mole fraction
of CrF,, together with the resulting equilibrium
quotient. In each case the partial pressure of
hydrogen was determined by subtracting the partial
pressure ‘of HF from 1 atm. The equilibrium
quotients are also shown graphically in Flg. 2.3.28
as o function of the CrF, content,

An examination of the results indicates: that,
within the experimental precision achieved, the
equilibrium quotients are independent of the mole

fraction of CrFé in the concentration range studied.

In order to obtain the activity coefficients of
CrF, in this solvent at 850°C, the following equi-
librium constants were calculated from available

106

tabulations ¢ of thermodynamic properties:

1. for CrF ,(s) + H,g) = Cr(s) + 2HF(g)

= 5.1 x 10-3 ;
2. for Cer(l) + Hy(g) ——>‘Cr(5) + QHF(g)

-—80x10' .

Comparison of these constants with the average
value that was obtained for the equilibrium ratio,
(1.45 £ 0.17) x 10-3, yielded values of activity
coefficients of CrF of 0.284 and of 0.181 at

 

16, Brewer et al,, NatlL Nuclear Erzergy Ser. Div. IV

198, 65, 109, 201 (1950).

+%

. iyt

 
 

 

 

PERIOD ENDING JANUARY 31, 1958

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-3 UNCLASSIFIED
(x 1077) . ORNL— LR—DWG 27937
4.0 l ‘ |
3.5 -
Ky =(1.452047)x10 ® DETERMINED UNDER REDUCING CONDITIONS
—~3 & DETERMINED UNDER OXIDIZING CONDITIONS
30 Ky(s)=51 x 10
Ko (/1 =18.0 x 1073
y{s}= 0.284
2.5 y(/}=0.181
%
o 2.0 " i
£.5
e e
1.0 A
0.5
0
1.0 2.0 3.0 4.0 5.0

6.0 7.0 8.0 9.0 (x1072

X
Cer

Fig. 2.3.28. Equilibrium Quotients at 850°C of the Reaction Cer(a') + Hz(g) = Cr(s) + 2HF(g) in NuF-Z‘rF‘.

(53-47 Mole %).

850°C with respect to the solid and the super-
cooled-liquid standard states, respectively, Ex-
perimental work is under way at other temperatures.

Solubility of F"eF.2 in LiF-BeF,
R. J. Sheil

Before attempting to determine activity coef-
ficients of FeF,

~tration  range below which the FeF2 does not
precipitate as @ ‘complex compound or in the pure -
For this mvestlgahon, 'known amounts offifi

~ Sstate.
_iron were added, as FeF,, to the solvent, ond
filtered samples of the solz

“added.
separafe oddrtlons of FeF,

Table 2.3.17. The tobulotecf results indicate that

-equ:llbrwm measurements can be made w:th -$0-
lutions of FeF,
well in excess of 5000 ppm Fe'*
which is the lowest temperature contemplated for
the experiment.

To_ble 2.3.17. Concentration of Iron Dissolved
in LiF-BeF2 {63-37 Mole %) ot 500, 600, and 700°C

 

Fe++ Found in Filtrate (ppm)

 

 

in molten  LiF-BeF, (63-37
" mole %), it was necessary to estabhsh a concen- -

Fe'' Added
(ppm) At700°C At 600°C At 500°C
5,000 4,925 4,835 4,845
10,000 9,386 9,125 9,390
141,700 33,100

* 50,000 49,100*

 

ution were obtained at =

_ different temperatures. The concentrations of iron - -
in the filtrates; as determined by chemical analysis, =~~~
‘were then compored with the ‘amounts of Fer":""

The results of the compartsons for these -

in this solvent ot concentranons' .
even at 500°C,

- *Pefi-ographie examination of this filtrate revealed that

._the sufurnhng phase was pure FeF2

Use of Crs ' To Study Cbromaum Mlgrofion in
Polythermul Inconel-Molten Sulf Systems

Ro Bt Evans -

. A proposed grophlcal method for colculatmg the -
‘are summarized in.  rate and amount “of : migration: (corrosion) to be
" expected

’ becouse of the reochon C

" Inconel<Na F-ZrF -UF

systems

Cr° 4 2UF, -.———2UF + CrF

has been evolved through studies of chromium
migration within polythermol loop systems., The

107

 
 

 

 

'MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

@alcu\qui‘ons are based on the assumption that the--
circulating salt initially contains equilibrium-

concentrations of UF4 UF,, and CrF,, which are

independent of time. In an attempt to compare the

calculated results with results of actual loop
experiments, a study was made of four thermal
loop experiments '7 in which Cr3! was utilized to

trace the chromium mlgrcmon patterns within the
loops. Several important points regarding the use

of radioactive tracers in loops were noted in the

course of this work and are described below, since

they will be of definite value in investigations of
corrosion in polythermal loop systems. An outline
of the experimental conditions for the four loop
tests is given in Table 2.3.18. Three of the four
experiments, as indicated in Table 2.3.18, were

conducted with Cr31 initially present only in the

salt, as Cr*F_, ‘The results of these first. three
‘experiments, in which Cr® was initially absent

- from the walls, do not give a direct indication of

the over-all chromium migration pattern as a
result of corrosion. To illustrate this point, a
combined plot of the experimental data for the
first three experiments of Table 2.3.18 is shown
in Fig. 2.3.29. If it is assumed that the loop wall
distribution patterns for Cr® and Cr coincide, it
could be concluded from Fig. 2.3.29 that the zone
of maximum deposition of chromium as a result of
the reversal of the corrosion reaction is near
section A. By mass-balance calculations, A is
between the zones of maximum chromium corrosion
or depletion, C-C’ and the zones of maximum

deposition as a result of corrosion, B-B% _In

-other words, the loop wall distribution patterns for

 

17R. B. Price et al.,, A Tracer Study of the Transport
of Chromium in Fluoride Fuel Systems, BMI-1194
(June 18, 1957).

Cr°® and Cr™ do not coincide when radicactive
chromium is initially mtroduced to the system as

,CrF

Corve 2 of Fig. 2.3._29.';s.hows,q secondary

- maximum near the region of the true deposition.

The original invesfigatérs” have suggested that
the shape of curve 2 is of greater significance
than are the shapes of curves 1 and 3, because the
loop operating time for the data of curve 2 was
400 hr as compared with 125 and 288 hr for curves
1 and 3, respectively. The salt .in loop No. 3

- was, however,. pre-equilibrated for 192 hr before

operation. It would seem that, if the secondary
maximum of curve 2 were truly significant, the
other two curves should have shown some tendency
toward the initial formation of such a maximum.
It is difficult to visualize a chromium transport

o, UNCLASSIFIED
B B AC €  ORNL-LR-DWG 27938

| [ o ! '
croINTIALLY PRESENT R |
IN SALT,BUTNOT IN / : . ]

WALLS sk

: W EXPERIMENT {
T=="1 \&‘/EXPERIMENTZ =

/ \\{,EXPERMENT 3

\ _

{
i
I
f
4 '
1 : 1
{1/ \
i
! .
/

¥
.
/ \
f — b
! / \\\
_—"‘7“ N \\
4 Sl
| .~ "-.__,__
i ] ] !
0 20 40 60 80 100 80 &0 40 20 ©
PERCENTAGE OF LOOP TEMPERATURE DIFFERENCE .
7 {min) 7 (max} 7-(min}

2
3

 

&4

PERCENTAGE OF MAXIMUM WALL ACTIVITY FROM Cr™"

 

 

 

@
o
N

 

 

»
o

 

 

-

H
o
.y
s‘

\-
o
o
=

n
Q
.

 

 

 

 

 

 

 

 

 

 

 

 

o

 

Fig. 2.3.29. Effect of UF -UF3 Corrosson on Cr51
Distribution in Inconel Loops. - : :

Table 2.3.18. Summary of Experimental Conditions of‘Bnt_tellle Thermal Loop Expe_rimenf‘s:. |

 

Expefimenf No.

 

 

1 2. 3 -4

' Cr5'1‘ initially'ifi the melt Yes Yes Yes | Yes

Ce31 initially in walls - | No No No o 7 Yes

UFs-und CrF, ifiitiully in melt ‘  Ne No. ~ Yes - o Yers

" Total op'elr@ting time, hr = - 125 400 288 L 288
Average Ioop.tempe.rcture, °F 1356

1341 ' 1352 : 1345

 

108

«V.

C

 
 

 

mechanism whereby the relative amount - of
chromium deposited at a point suddenly jumps
from a constant valuve of 30% of the ‘maximum to a
valve of 75% of the maximum after 288 hr of
operation. Calculated Cr31 distribution curves for
this type of experiment under conditions of perfect
pre-equilibration of the -salt have the shape
exhibited by curves 1 and 3. They also have a

constant shape with respect to time when plotted -

on a relative basis. The theory on which these
calculations were based is outlined below.

The Cr% distribution under any experimental
procedure is governed by two chemical equilibria
at the wall surface, the most important of the two
being the wall exchange reaction

Cr° + Cr*F2'¢— Cro* + CrF2
for which |
NCrF
(1) K, =
NCr IYCr*F 2

i

where N denotes mole fraction of the material
indicated by the subscript. Equation 1 is inde-
pendent of temperature. The other reaction that
affects the Cr®* distribution is

Cr® + 2UF <— 2UF, + CrF

in which the chromium may be labeled or unlabeled.

Thus

S NUF N(CrF + Cr*F2) o
@ .Kzf-= e o
- NUF‘ N{Cr + c#’*;

 

~is zero at these points,

PERIOD ENDING JANUARY 31, 1958

The net chromium migration across the wall surface’
Accordingly the wall-
surfoce chromium concentration at the balance
points is- equal to the original chromium concen-
tration in the alloy.

In addition to the chemical equilibria at the
surface, the Cr%* (and Cr® wall distribution is
governed by the concentration of Cr®* (and Cr°)
initially in the interior of the walls as a result of
the solid-state diffusion phenomena that are
characteristic of metals at high temperature,
Since linear diffusion in a semi-infinite medium is
involved, the distributions also depend on the
square root of the -diffusion coefficient, D, for
chromium in Inconel. The diffusion coefficient is
practically the same for Cr® and for Cr%, and the
value is very sensitive to temperature changes.

A combination numerical-schematic steady-
state example similar to conditions of the third
experiment and involving the various chromium
concentrations in three loop sections is presented
in Fig. 2.3.30. Equations 1 and 3 were used to
prepare the example. It was assumed that, at
a gwen instant of time,

Nepep, = 1 X 10-10 ,
NCI‘O E 00]74 7
Tzp = 1010°K
[hus Nic,p scvr,) = 252 1073, independent
of timel, and the maximum salt temperature is

1042°K (T edn : = 997°K). The _salt was assumed

o “to be NaF- ZrF, UF (50-46—4 mole %, initially).
o [thus N

(CrF2+c.*F2) 2 52 % 10"73, independent of

. ___ji-_'rrflme], and the maximum salt temperuture is 1042°K

'vThe equ;llbrlum ratio K s femperqfure dependent.'}-l]
- i the UF UF ., 3 " and CrF concentrations in- the

V,?,L qs

[N(Cr + Cro*) (Kz)]

 

'(3) ce ° % cf"*)L (K

2

deposition and depletion. In a properly pre-

equilibrated loop these points remain stationary.

T

- all chromium,
T ‘Iubeled chromlum. w

- = 997°K)

mban

of _unlabeled chrommm, and of

1s posmve.

The ‘most sugmf:cont pomt demonsfrated by

" Flg. 2.3.30 is thot calculated instantaneous Cro*

loop wall distribution data that tcke corrosion
into account will give approximately the same

109

The sclt was. assumed to.be
] NaF-ZrF -UF4 (50-46-4 mole %, lnmaHy) The

._ salt are- constant with reSpect to time, Eq. 2 may |
- "be -rewritten as .a. functton of any loop posmon, "‘quunhhes (dM/dt)T, (dM/d;) e r‘°’ “and. (dM/dt) oy

represent, respechvely, the net - diffusion rate “of

_ The rates aré positive when
8 _;:_-r‘the flow occurs ina posmve x direction, that is,
_ , S e G '-f_:.f.'when [Cr°] <=0 -[Cr] ‘
The subscnpf BP refers to balonce pomts, whsch;,:’
are the two points  that separate the zones of -

 
 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED
* ORNL-~ LR- DWG 27939

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a1 TRan | X400 Cop, | Coxr, | Goo | Coox. | BGG0 ACgox | Coo 1 Coox
1K) - .CONCENTRAT_ION, Cimole %)
. — " /,r”” = T
1042 : {042. 332 /0/2/5 2 /5//‘9/// +63X10 i i
. : -dM.) - (dM). : (dM)« o
- amy - A oy (M >4
‘ = o (-d’ 7+« dat Cr0 < It Jeox o
967 | olo | 303 [/0.2527¢ 14,{{6”3': t+6.9xi0TH
I aM : ’
! (d}') Q o \er
: - ! . 77777 STTTIT V
‘952 952 | 252 /025 A 1x10-87
. ) i _ //////
; am -
l (df) >1
l T —> .
o ¥ = X
(.2m & - , |
(—;,—) = NET DIFFUSION RATE OF ALL B
E r CHROMIUM | - -
|/ ,
. % MELT IN CONTACT WITH WALL.
(—d—,—) = NET DIFFUSION RATE QF UNLABELED S ,
ce® CHROMIUM SURFACE LAYER ASSUMED TO BE

- \q? .= NET DIFFUSION RATE OF LABELED
’ Cr°* . CHROMIUM

~Fig. 2.3.30.
Loop Wall Segments, .

curves as those obtained experimentally (Flg.

2.3.29) because (1) Cr®* is always diffusing into
the walls along the entire loop, regardless of the
sign or value of the other flow rates, and (2) the
Cr® diffusion rate is highest in the high-temper-
ature region,. since (dM/dt)Cro, is proportional to
(AC o,oD‘/z) and D172 jncreases with temper-
ature. It may be concluded that experiments
similar to the first three listed in Table 2,3.18

are of little value for corrosion studies because =

~ the Cr% distribution pattern is controlled by the

wall exchange  reaction and wall temperature
- rather than by the corrosion reaction of interest.
In the fourth experiment, Cr5! was mmally
present both in the salt, as Cr*F o and in.the
walls, as Cr°, The labeled chromium migration
“pattern in this case tended to follow the over-all
" migration poflern of the . corrosion, as shown
grcphlculiy on  Fig. 2 3. 3] Comparisons  of

~at the inner surface of the tubing is very close
_to the salt temperoture and that this temperature
controls * the value of the ethbnum constant

IN EQUILIBRIUM WITH MELT

 

i1 INFINITE INCONEL TUBE

Theoretical Concentration Relationships for Chrcmlum in Vurlous Inconel Thermul-Convechon

'.Figfi.: 2.3.31 and 2.3.32 - demonstrate that the

- corrosion pattern calculated by assuming that

T, att)x=g = Tyqy is in fair agreement with the
pattern -obtained experimentally. -
mental curve of Fig. 2.3.31 is not symmetrlcal

- and the point of maximum deposition appears to be
. displaced downstream - from the maximum wall
‘_femperature at ‘section A.-

features  characterize the curve of Fig. 2.3.32.

This strongly suggests that the wall temperature

' for the corrosion reuchon.

The ability toreproduce the shape of the activity

vs wall-temperature plots for the Battelle thermal

loop experiments — particularly for the fourth

‘experiment - by means of calculations based on
- the predlcted behavior suggests that the proposed

The experi- -

In general, the same

 
 

O

 

- of !nconel. i iy

PERIOD ENDING JANUARY 31, 1958

UNCLASSIFIED
ORNL-LR—-DWG 27940

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

f'mlgratlon mechamsms gwe a reasonable expla-'_‘-fil.?-'
' nation of  the - steady-state chromwm mlgratlon
"'wnhtn thermai I°°Ps",;-,r.~' . s T

Actuvrty of Nrckel m NlekeI-Molybdenum Alloys ‘_>;_ : 2. mechamcal ag|tat|on has no . effect on- the

3. Longer

Measurements are bemg made of the actwnty ef
nrckel in mckel-moiybcfenum u”oys by using an
-The: rehab:llty af
T -',47, 138 (]951)

electromotive - force method.
easurements made wnth emf cells of the type

NlCl'z(] wt ’%) in

~ Ni(metal) AR
NaCl-KCl eutectic

NI-MO alloy

Ly . B’ A c!
2 160 —
Z T 1 ~
: | o
l
L
|8
& 150 l
~
Z 140 5 1,
- a 4100 |- WALL ™=, -
3 g 'r.
- - SALT
2 % 50 / el
S 120 2 \/ / . \
O ) o .
o O 25 50 715 100 125
§ % OF LOOP LENGTH
% 100 F——x% —
= — N - FLOW /,/ ——
3 / \\ <
L - /
O \
2 ao 1 A A
s - /7
5 - /
o : : '
7 N\ \ /
q [ curee - THERMAL ASSUMPTION —|" /
O - \ /
g Twarr] = TeaLt \ .
- g 0 mee——= Toarr] = TwaLL] -~ ¥ 7 _ / .
= ‘X =0 DENOTES INTERNAL TUBING SURFACE ]
LJ | -X—=o DENOTES EXTERNAL TUBING SURFACE \\l / .
g' 20 — ; - 7
8 \ /
L) .
= .
.2 .
: ] \
5 1 -
& 0O . -
5 0 20 - 40 60 80 100 BO 60 40 20 0
_ PERCENTAGE OF LOOP TEMPERATURE DIFFERENCE
T(mm) S ‘ T(max) : ‘ T(min)
Flg. _23 31 Calculated Loop Wali Dlstribut:on af Chromium After Diffusion-Controlled UF -UF Corrosion

_f;'.-ls dependent upon the reversrblhty of the ce!l . |
"The “usual criteria18-20_ accepted as evidence

'~ that the cell is’ ‘behaving reversibly are that
Yithe: potentcal remains, constant w:th time at -

;-.flconstant temperature, '

G '5_' j;k—i’potentral 7
3. the . potential - of the cell returns to its “equi-

""”f:ilubrlum value rapldly after palarlzatlon,

184 F, Elllot nnd J Chipman. TranS- Fafaday Soc.

: 9K W, Wagner, Tbermodynamzcs of Alloys, p 921 —97
Addison-Wesley, Caombridge, 1952.

2G5, F. A. Kortum and J. O'M Bockris, Textbook of
Electrecbem:stry. p 581, Elsewer, New York, 1951,

m

 
 

 

 

 

MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT

UNCLASSIFIED

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘ B A C ORNL~LR-DWG 27941
120 : ] |
3 100 _ e
w
o< ® e
gg | p . FLOW
& . 80 ) \\
. ZI._L [
O > \ @
EE | | .
aZ S ® ¢
ok 60 o -
< : ®
tn
= |
33
=
= 40 : /
zs 1 ¢
o2 EXPERIMENT 4: Cr INITIALLY
_-n‘_)?z PRESENT IN WALL AND IN SALT _
g E \ o
Z 20 .‘ /‘
. 3
0 | &L
0. 20 40 60 80 {00 80 60 40 20 0
: PERCENTAGE OF LOOP TEMPERATURE DIFFERENCE ,
T(mvin) 7 (max) 7 (min)

Fig. 2.3.32. Effect of UF4-UI"'3 Corrosion on Cr° | Distribution in Inconel Loop.

. 4, the observed potentials are reversible with
_ . respect to temperature changes,

5. there is no difference of potential between
two apparently identical electrodes.

The first condition has been met by all cells
tested. Also, mechanical agitation of the elec-
trodes has had no effect on the observed emf. The
cells have also returned to their equilibrium
potentials within 30 min ofter all four electrodes

have been shorted together. In most cases, the

electrodes are within 2 mv of their equilibrium
potential within 5 min after shorting for as long
as 15 min. It was found that concentration cells
in the nickel-chromium system were not readily
~ reversible with respect to temperature changes.?!
" However, the first attempt to run a cell at two
different temperatures in this study was suc-
cessfuls A cell with alloy electrodes containing

 

2Vyork done by M: 8. Panish, ORNL.

112

10 oand 20 ot. % molybdenum gave constant po-
tentials at 800°C for 15 days and for 6 more days
at 700°C. It is planned, in the future, to ascend
and descend the temperature scale between 750
and 1000°C with all cells. .
The criterion of reversibility, however, has been
the cause of the early termination of several cell
experiments. Each cell as presently run contains
two alloy electrodes (usually of different compo-
sition) and two pure nickel electrodes that were
annealed at 2200°F for 16 hr after fabrication.
in general, one nickel electrode becomes erratic
either shortly after the start of a run or after
behaving properly for as long as 12 days.
Potentials as high as 70 mv have been observed

between two nickel electrodes which two days:
previously 'had been behaving properly ond

exhibited no potential difference. = Cells con-
taining four nickel electrodes from different
sources are now being run in an effort to discover

the cause -of this anomalous behavier, No
explanation is evident to date. S

 

 
 

i

 

 

contained BeF

PRODUCTION OF PURIFIED MIXTURES
Jq Po Blakely Gl J- Ness‘e
Preparation of Pure Fluorides

Some transition-metal fluorides of high purity
were prepared for use in studies of corrosion,
chemical equilibria, and phase relationships,
ond for use as reactants in the preparation of
other flucrides, including complexes. Nearly a
kilogram each of ferric, ferrous, nickelous, and

cobaltous fluorides were prepared. Ferric fluoride

was prepared by passing anhydrous hydrogen
fluoride gas over anhydrous ferric chloride at about
300°C. For preparing the ferrous, nickelous, and
cobaltous fluorides, the commercially available
hydrated chlorides were partially dehydrated at
about 100°C and treated with hydrogen fluoride
at about 400°C,

Small+Scale Purification Operations
C. R. Croft
The experimental facilities were used for the

processing of 530 kg of salt mixtures in 55 batches
that consisted of 5-, 10-, and 50-lb quantities..

Twenty-eight of these batches were materials that
g+ Included in the total were 11
batches of the NoF-KF-LiF-UF, mixture for use

by the Metallurgy Division.

Requests for molten ‘salts and especrally for
BeF ,-bearing mixtures have reached a level which
is beyond the production capabilities of the
experimental facilities. It is anticipated that the

250-1b-batch production equipment will be operated
at intervals during the ‘next severul months to
e 'prowde the material reqmred ;

“Four. batches totaling 270 kg of LIF-Ber-UF‘

- mixture were prepared during a trial run of thef'
~.250-1b- batch production equipment for two weeks
: durmg December. This operation, on a three- ,
o shift, flve-day-week b05|s,' was - conducted - o
.~ cooperation ~ with the ~Y-12 industrial Hyg:enel_
':"'Deportmenr, and face. masks and complete pro-
" tective - clothlng were - used by the operating
" personnel, Careful monitoring by the ‘industrial
- hygienists-: reveuled no,_significant exposure of the -
- personnel to air-borne berylilum compounds.r The
.- occurrence . of . severe “dermatitis among the
operators, however, indicated that routine operation

of the equipment under the conditions of this test

PERIOD ENDING JANUARY 31, 1958

was not feasible. It will, apparently, be necessary
to enclose the equipment and to provide an
improved ventilation system before its continued
use can be odjudged saofe in all procticable
circumstances.

Anclyses indicated only 3 to 4 wt % uranium in
the final product, rather than the 6.4 wt % charged
to the reactor. ' Examination of the residue in the
reactor showed high concentrations of uranium as
UO,. It is apparent that contact between HF
and the liquid in the large reactor was insufficient
to remove H,O and BeO from the mixture and that
precipitation of UO, by the reaction

UF, + 2BeO —>UO, + 2BeF,

resulted. Experiments are under way to develop
a modified technique for preparing satisfactory
material with this equipment. Most of the available
material con be used in experiments for which the
precise compositions requested are not essential.

-Preparation of Material for InsFile Loop
- F. A. Doss
A single batch was prepared that totaled about

300 g of LiF-BeF ,-UF , mixture in which Li7 and
U235 were used. The solvent mixture of Li’F

and BeF, was purified in Building 9928 at the
Y-12 Plant. When analyses of specimens from

“this preparation had indicated its purity to be
“satisfactory, the material, in its original con-

tainer, was taken to Building 9212 at the Y-12
Plant where the U235F was added and the final

_purification was occompllshed Chemical analyses
. .made by Y-12 and ORNL laborutorles were in good
 agreement as ‘fo. composmon .of the melt for
'_7accountab||sty purposes. . The material has, -
~accordingly, been trunsferred to the requestor (see
Lo Chap. 2.2 of thls report)

Trunsfer and Servrce Gperatlons
F. A Doss -

Flfty-flve flllmg and . dramlng ‘operations were

':'.',performed during the quarter. Requests for such
“services declined to a low of 12 during December
‘and -then increased to 25 during January. About

640 kg of salt and about 300 kg of alkali metals

were transferred in these operations.

113

 
 

 

 

 
 

10
1.
12.
13.
14,
135.

- 16.
17.
18.
19.
20.
21.
22.
23.
24,
25.
26.
27.
28.
29.
30.

- 31-33.
Y
35,

e
o
39

 

 

S A
42,
S s,

 

\\
VOENO LA LN

a0

. Alexander

. Bettis

. Billington

. Blakely

. Biankenship
lizard

. B. Briggs

. 0. Campbell

. W. Cardwell

. H. Carr

. |l. Cathers
E. Center (K-25)
A. Charpie

J. H. Coobs

J
F
E.
A
C
., G
M.
E. J. Breeding
R
D
D
W
G
C.
R.

'F. L. Culler

J- Ho DeVan

L. B. Emlet (K-25)

V. K. Ergen
J. Y. Estabrook

'D. E. Ferguson

A. P. Fraas

E. A. Franco-Ferreira
J. H. Frye, Jr.

A. T. Gresky

J. L. Gregg

W. R. Grimes

E. Guth -

- "C. S. Harrill
36,0 |

.A. Hollaender -~ - =

-A.S. Householder s

< W.H.Jordan . =

G. W, Keilholtz

H. W. :Hoffmuh

C.P.Keim

M. T, Kel!ey 7
F. Kerfesz o |
B. W, Kmyon *_ e

M E. Lackey

INTERNAL DISTRIBUTION

46.

47.

48.
49.

50.

51.

52.

33.

4.
55.
56.
57.
58.
59.
60.
61.
62.
é3.
64.
6é5.
66.
7.
é8.
&9.
70,
71.
72.
73.
74.
78.
76.

77.

78,

79,

80.
8l

82
83,

. .100.-
101 ]03._

EXTERNAL DISTRIBUTION o

104. Division of Research and Development, AEC, ORO
v 105-681. Given distribution as shown in T1D-4500 (13th ed. Rev.) under Reactors-Power category
’ . (75 copies = OTS) ' :

ORNL-2474
UC-81 = Reactors=Power
TID-4500 (13th ed., Rev.)

February 15, 1958

vingston
ac Pherson

. Morgan

. Murray (Y-12)
L Nelson
. Patriarca
A. M. Perry
D. Phillips.
P. M. Reyling
J. T. Roberts
M. T. Robinsen
H. W. Savage
A. W. Savolainen
J. L. Scott
E. D. Shipley
M. J. Skinner
A. H. Snell
J. A. Swartout
A. Taboada
E. H. Tayler
R. E. Thoma
F. C. VonderLage
G. M. Watson
A. M. Weinberg :" '
M. E. Whatley

_G. D. Whitman
- G. C. Williams
.C. E. VWinters
~Jo Zasler

ORNL Y-12 Technical Librury
Document Reference Section
Laboratory Records Department -
Laboratory Records, ORNL-RC
Cenfrd | Research Library

115

 
 

r

 

 
 

 

S

)

 

ORNL.-2474 - _
UC-81 -~ Reactors—-Power

Contract No. W-7405-eng-26
MOLTEN-SALT REACTOR PROGRAM

QUARTERLY PROGRESS REPORT

For Period Ending January 31, 1958

H. G. MacPherson, Program Director

DATE ISSUED

lwlfiJ

MAY 141958

 

e omc RIDGE NATIONAL LABORATORY -

- Qak Ridge, Tennessee -
-operated by
UNION CARBIDE CORPORATION
for the

U.S. ATOMIC ENERGY COMMISSION

 
 

 

 

g 1ok e

"