AR

3 4456 0360719 4

 

 

- M e R 8 Y i

CENTRAL RESEARCH LIBRARY
DOCUMENT COLLECTION

i
i
F LIBRARY LOAN COPY
'f

    

DO NOT TRANSFER TO ANOTHER PERSON

If you wish someone else to see this document, ."

send in name with document and the library will
arrange a loan.

 
 
 

 

ORNL-1965

This document consists of 56 pages.

Copy}/f of 232 copies. Series A,

Contract No. W-7405-eng«26

SOLID STATE DIVISION

A FLUORIDE FUEL IN-PILE LOOP EXPERIMENT

O. Sisman, W. E. Brundage, and W. W. Parkinson

Work Done by

C. D. Baumann W. W. Parkinson
W. E. Brundage O. Sisman
R. M. Carroll F. M. Blacksher
J. G. Morgan C. Ellis
M. T. Morgan J. R. Duckworth
A. S. Olson

DATE ISSUED

JAN 151357

OAK RIDGE NATIONAL LABORATORY
Operated by
UNION CARBIDE NUCLEAR COMPANY
A Division of Union Corbide and Carbon Cerporation
Post Office Box X
Qck Ridge, Tennessee

 

MARTIN MAR STEMS LIBRARIE

AR

3 4456 0360719 Y

 

 

 

 

 

 

 

 

 

 

 

 
VNG ARWN~

. G. Affel

. J. Barton

. Bender
Billington
. Blankenship
. Blizard
Borkowski
. Boudreau
Boyd

. Bredig

. Browning
. Bruce

. Cadllihan

. Cardwell

. Center (K-25)
. Charpie

. Clifford

. Coobs

. Cottrell

. Cowen
romer
Crouse

. Culler

. DeVan

. Doney

. Douglas

. Dytko

. Eister

. Emlet (K-25)
. Ferguson
. Fraas

. Frye

. Furgerson
Breeding
. Grimes

. Hoffman

. Hoffman
olfaender
Householder
. Howe

. Jordan

. Keilholtz
. Keim

. Kelley

OUEPIMPMEOAOMPMME="TNY

AU E I AOVIEMIO—ATIMOPXRXIOPZIIC®D

. Kertesz

. M. King

MMEONEE-PPIMEME-POrEMOrE-MAVODE-ODOIPNEZOEOMMOZIOD

 

 

INTERNAL DISTRIBUTION

47.
48.
49.
50.
51
52.
53.
54.
55.
56.
57.
58.
59.
60.
61,
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.

75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
1.
92.

ORNL-1965
C-84 — Reactors-Special Features of Aircraft Reoctors

CAMEOPOECOPMAIPIZPIICPRAIECPOELMACIPPOMECMENDDD L

D= mMmTTnNIx=x

Si

TEODCrrITENEMET<ADPOONZOD >

omI» O T V-

. ane
. Lindaver

Livingston

. Lyon

. Maienschein
. Manly

. Mann

. Mann

. McDonald

. McQuilkin

. Meghreblian
. Milford
Miller

. Moore
. Morgan

. Morgan
Murphy

. Murray (Y-12)
. Nelson
Nessle

. Oliver

. Overholser
. Parkinson
atriarca

. Peelle

. Perry

. Pigg

. Poppendiek
. Reyling

. Richt

. Robinson

. Savage

. Savolainen
. Schultheiss
. Shipley

. Simon

. Si

sman

tes
Skinner

. Smith

. Snell

. Susano

. Swartout

. Taylor

. Thoma

. Trauger

 

 
 

 

 

 
 

 

CONTENTS
A B ST R A C T ettt es v e st st asbe s et aan e e este st e st hsbassbs St ees en et en e S a st enn e b s A s e e b e R e e e e e ra e e 1
INTRODUGCTION e ettt et oo e eeee e et estestestsetessesabessansseese saseseaeaeseeeeeasassubasoRessa e s e s b e sa e s Ren e b e ab e s s s eneabasbes bt sbean et ne 1
SUMMARY oottt ettt e e e e et ebetaees s e st ebeeas et e s araas e st aseee e seemteateatbebesas e b b e SR e on s Sensae eRe L Ra et et e st st 2
DESIGN ottt te ettt ee et e st eeesebesa b e ssesae e basseseaestessesnaese s s e beb bbb e s b e R Ee e S AR SRR e ST e Ae RS e Re e Rt e nA e E e e 3
GOVEIMING FACTOFS ittt bbb s b e et bbb b 3
Experience with the First Loop ..o e 5
The Basic LOOp APPAratus ......coieiivieiiieenineiscieresiie it ieeai bt sa e bbb e sb e 6
Safety ............................................................................................................................................................ 6
The AUXilIAry Sy StEMS .oueiiirireeiieee ettt bbbt e e bbb 8
CONSTRUGCTION oottt it oottt ee s e e e e ae e e s e ase e ssaat e st e beassos st s e s e e s e s e e b 4 obs e b ab e s s s beer e s b st e et sa b 8
T @ U QU coeeeeeeeeee et ettt ettt cettesssabeseseeaebesbees s se e s bessse e s ee b e s oate s s it e s e s en s e e it abe s ba s s abn e see s e be s st bt s e as 8
The PUMP SYSTEM .viviivieieeeeeirieceststeeaeseieie s eratbe s ers st s s h s s s b s et h bbbt bbb 11
The Off-Gas SySTEm ..ottt eb bbb b e s 12
The Assembly Procedure..... ..ottt 15
Instrumentation ANd CONTIOl ..ottt se et b s b s bbb s 18
Radiation Shielding .....civvereiereiet it e 22
P E R ATION oottt e ete st e et e e e e s teseesbe et easaas s e s aatesheeseemtsaber e s se e oRE e s e ek b e b s e b e AR e AR e e R e st st e b e 26
Leak Testing and Cleaning......cccoriiiiiiniieie sttt s 26
FS Al I QEION 1M REGCEOT caenveeee e et tee et ee e et esteate s etaebesasesee e st sonaesassan e e be s sheaas e e ba s s s onres s s meend s b b s s an s a e s 27
F LT ING eevetiteeeee et b et enee st e e s e st bbb e b b e b e e e AR SEAELehERES LSRR S e 27
Reactor Startup and Operation of the Loop ..o 27
ShutdOWN Of the LOOP .vevee ettt e st 30
Removal and Disassembly of the L.oop ..o 31
DISCUSSION OF RESULTS oot ceeeteesbessesneseesbasre s eesasas esbaesessasasaabesassssaasess e sabe bt easn e suasr b s s nrn e e 32
Measurement of Fission Power Generation ...t 32
Chemical Analysis of FUl .ttt e 33
FisSiON-Product RUMENTUM c.oeeeee et eette e et e e st e bar e e ee s ns s as s s n s e s b e aaa s s e s b s as s s bbb 34
Metallographic EXGMiNGtion ..o es ettt st 34
CONC USTOMS 1ot eeeeeeeee e esetee e e ee e eeeaeeeaate s asbeaeates s essaeeseaaee e s aee s saesenbbe e abn e s s b e e e s n s e bm st e et e e s e ne s n b e bt v s 45
APPENDIX A —~ DISASSEMBLY PROCEDURE AND EQUIPMENT ..o 45

APPENDIX B — ANALYSIS OF INCONEL TUBING ... vttt a e et aas 50

 

 
 

A FLLUORIDE FUEL IN-PILE LOOP EXPERIMENT

0. Sisman W. E. Brundage W. W. Parkinson

ABSTRACT

An Inconel loop circulating fluoride fuel (62]/2 mole % NaF, ]2]/2 mole % ZrF4, 25 mole % UF4,
93% enriched) was operated at 1485°F with a temperature difference of about 35°F in the Low-
Intensity Test Reactor for 645 hr. For 475 hr of this time the reactor was at full power, and
fission power generation in the loop was 2.7 kw, with a maximum power density of 0.4 kw/cc.
The tatal volume of fuel was 1290 cc (5.0 kg), and the flow through the irradiated section was
8.6 fps (Reynolds number 5500).

chemical and metallographic analyses.

The loop has been disassembled and has been examined by
No acceleration of corrosion or decomposition of fuel by
irradiation was noted, although deposition of fission-product ruthenium was observed. No mass
transfer of Inconel was found, and the corrosive attack was general and relatively light, The
average corrosive penetration, in the usual form of subsurface voids, was 0.5 mil; the maximum

penetration was 2 to 3 mils.

 

INTRODUCTION

High power densities and effective heat transfer
are paramount requirements in the development of
mobile power reactors. In order to provide a system
meeting these requirements for the Aircraft Nuclear
Propulsion Project, circulating molten-salt fuels
have been investigated extensively, This experi-
ment, as the first of its kind, was designed and
operated primarily to demonstrate feasibility as
well as to study the effect of reactor radiation on
a dynamic-corrosion system of high-melting liquid
fuel. It was anticipated that the fluoride fuels
would be used at 1500°F in Inconel.

At the time of the conception of this experiment,

1 for several

there were available corrosion data
alloys in a number of compositions of fluoride
salts at temperatures up to about 1800°F. There
were also available data for very small Inconel
capsules containing the fluoride mixtures which
had been irradiated in the MTR at about 1500°F.
Very little, if any, corrosion of Inconel was de-
tected for either the irradiated or the nonirradiated

capsules of the more recent tests.2:3  Thermal-

 

1. S. Richardson, D. C. Vreeland, and W, D. Manly,
Corrosion
17, 1953).

2y, J. Sturm, R. J. Jones, and M. J. Feldman, The
Stability of Several Inconel-UF4 Fused Salt Fuel Sys-
tems Under Proton Bombardment, ORNL-1530 (June
19, 1953).

by Molten Fluorides, ORNL-1491 (March

 

convection loops of very low flows but large
differences in (AT)

transfer to be a serious problem.4 No high-velocity

temperature showed mass
experiments had been run either in the reactor or
out of the reactor. It was anticipated that the
Aircraft Reactor Experiment would be the first
high-velocity system, either with or without radi-
ation. Here, the flow rate and the AT would have
values expected for a full-scale reactor, but the
power density would be too low for an accurate
measurement of any effect of radiation. A loop
experiment is different from the ARE in that a
part of the system can be in a high-radiation field.

Loop experiments are therefore designed to
produce a very high power density in some part
of the system and to maintain, as nearly as
possible, many of the other conditions expected in
the reactor. ldeally, it would be desirable to have
a high power density and a low dilution factor
(ratio of total volume to irradiated volume), high
flow rate with high AT, and precisely the same
fuel composition as that anticipated for the full-

scale reactor.

 

3W’. E. Browning, Solid State Semiann. Prog. Rep.
Aug. 30, 1954, ORNL-1762, p 39.

4(3. M. Adamson, ANP Quar. Prog. Rep. June 10,
1954, ORNL-1729, p 72.

 
 

The radiation facility available for the experi-
ment was a horizontal beam hole (HB-2) in the
LITR. In order to obtain the highest possible
power in this reactor a special fuel was blended
which contained about 50 wt % U235, This mixture
of NaF, ZrF,, and UF, was somewhat harder to
work with than the standard mixture because it
had a higher melting point (1170°F vs about 950°F
for the ARE fuel) and a considerably higher
viscosity,

The intention was that the fuel loop be as short
as possible in order to keep the dilution factor
down and the fuel inventory low, The first design
had the entire loop located in the innermost 4 ft of
the beam hole, and this had the additional obvious
advantage of permitting the use of equipment
which was relatively small and, being completely
inside the reactor, was easy to shield. However,
this design required the development of a pump
small enough to fit inside the beam hole of the
reactor, Unfortunately, the urgency of the experi-
ment would not allow for the development of a
special pump. The only readily available one
was a sump pump, which had to be located outside
the reactor. As will be discussed later, this
changed the entire conception of the experiment,

The loop now became 15 ft long, contained a large
inventory of fuel and, consequently, a large
dilution factor, and required a tremendous amount
of radiation shielding outside the reactor,

With such a large dilution factor, the experiment
can no longer be considered a test of fuel stability,
but the effect of radiation on corrosion can still
be seen at the *‘in-pile’’ end of the loop, where the
radiation intensity is high, To simulate reactor
conditions it was desired that, in addition to
maintaining the maximum temperature of 1500°F,
a large temperature drop be maintained between the
hot and cold ends of the loop. Although the
attainment of a high AT was somewhat |imited by
the high melting point of the fuel, it turned out
that, because of the geometry imposed on the
experiment (see the section entitled ‘‘Design’’),
there was a choice between a high AT with fow
flow rate or a very low AT with high flow rate,
Since it was necessary and desirable to have a
relatively high flow rate, the loop was operated
practically isothermally, The temperature was
maintained at a maximum of 1500°F at the in-pile
end of the loop, and the flow was selected so
that turbulence was maintained in the hot end of

the loop (Reynolds number, ~5000).

 

SUMMARY

The experiment ran very smoothly throughout the
month that it was in the reactor. An early plugging
in one of the off-gas lines was easily corrected,
and the only adjustments required were heater
power reguiation for reactor shutdown and startup.
Temperatures, fuel flow rate, and reactor power are
plotted in Fig. 24 for the entire period of operation;
the conditions of operation are summarized in
Table 1. Termination of the experiment was
caused by flow stoppage due to failure of the pump
drive belt, The flow stoppage caused an overshoot
in temperature which scrammed the reactor by
means of the safety circuit built into the equipment,
The temperature overshoot was 70°F and lasted
for about 1 sec, after which time the reactor was
shut down (see Fig. 25).

Metallographic examination showed that cor-
rosion in the loop was relatively light and es-
sentially uniform, Maximum penetration was
2.5 mils, and such penetration was confined to a

limited region outside the radiation field, In the
remainder of the loop, maximum penetration was
about 1 mil, and average penetration was 0.5 mil,
No enhancement of corrosion by irradiation was
observed, and no mass transfer of metal was noted
anywhere in the loop.

Chemical analysis indicated that the fuel was
diluted by the preliminary flushing salt (nonuranium)
to 43.7 wt % (~25 mole %). The other changes
observed in the composition were in the traces of
nickel and chromium, constituents of Inconel,
The nickel content decreased slightly, while the
chromium content increased from 44 ppm in the
original charging material to about 150 ppm. These
changes confirmed the moderate degree of cor-
rosion,

The generation of power in the loop by fission
was estimated by electrical heating measurements
during operation to be 2.8 kw. Activation of the

Inconel of the loop corresponded to a flux which

 

 
 

Table 1. Summary of Operating Conditions

Totol operating time, hr
Time at full reactor power, hr
Location of experiment

Reactor po wer, Mw

645

475

LITR hole HB-2
3

Fission power in loop, kw 2.7
Maximum power density, kw/cc 0.4
Temperature, average, °F
Inlet to high-flux region 1475 + 10
Outlet from high-flux region 1485 + 10
Maximum in loop 1485 + 10
Minimum in loop 1450 + 15
AT in loop, °F 35 x15
Flow, averoge
Mass flow, g/ sec 250
Volume flow, gpm 11

Velocity in high-flux region, fps 8.6

R in high-flux region 5500

Velocity in most of loap, fps 2.7
Pressure (at pump reservoir) Atmospheric

Cycle time, sec

Complete circuit 19.6
Time in pump 8.5
Time in high-velocity section 0.4
Time in rest of loop 10.7

Fuel (as received)

Table 1 (continued)

Density ot 1500°F, g/cc 3.93
Viscosity at 1500°F, centipoises  10.0
Uranium enrichment, % 93
Container material Inconel .
Fuel inventory, kg 5.0
Volume of system ot 1500°F, cc 1290
Dilution factor 180

would have produced 2.8 kw. The maximum power
density in Table 1 was determined by this method,
The activity of zirconium fission product in the
fuel indicated that the power was about 2.4 kw,

Radiation instruments
activity of effluent gas from several locations
were included in the instrumentation as safety
devices, These instruments were affected by the
general background radiation but showed the only
other activity to be that from fission products in
the pump purge gas. The activity of this gas
after passage through the charcoal adsorber can
be estimated roughly by means of the stack-gas
monitor of the Reactor Operations Department,
About (.001 curie/sec was discharged to the
off-gas stack,

Radiochemical examination of the fuel showed

indicating only relative

 

Composition, mole % UF ¢ ZrF ¢NoF that less than 20% of the ruthenium fission product

(25-12'/2-62]/2) was retained in the fuel; the rest was found to be

Melting point, °F 1170 deposited on the inconel walls of the loop system,
DESIGN

Governing Factors

As mentioned in the Introduction, time would
not permit the development of a pump small enough
to be located inside the beam hole and equal to
the severe temperature and corrosion conditions of
the loop. The only suitable existing pump was a
centrifugal sump pump developed by the Experi-
mental Engineering Section of the Aircraft Reactor
Engineering Division., The capacity of the pump
was 600 cc, the output at 3000 rpm was about
1.3 gpm of fuel at 45 psi discharge pressure, and
the over-all dimensions were 23 in, in height and
10 in. The characteristics of the
pump dictated many of the features of the loop.

in diameter,

The beam hole in the LITR available for the

experiment extended about 12 ft through the reactor

shield and moderator structure and into the active
lattice. Since the diameter of the hole required
that the pump be located outside the reactor shield
at a distance sufficient to permit utilities con-
nections and shielding, the loop had to be about
15 ft long. A schematic drawing of the loop
apparatus in its final form is shown in Fig. 1.

The need to attain maximum fission-power
density in the fuel and the practical limitation
on the removal of heat, while an assembly was
retained which could be kept hot in the absence of
fission heat, required that the volume of fuel in
the neutron flux be small, The low thermal
conductivity® of the fuel made it mandatory to have

 

3S. |. Cohen et al., Physical Property Charts [or
Some Reactor Fuels, Coolants, and Miscellaneous

Materials, ORNL CF-54-6-188 (June 21, 1954),

 
- ~JACKET ASSEMBLY

    

NOSE PIECE

JLN HEATER
L

Al
®

FUEL TUBING

- NOSE COVER

ALSIMAG 202
HEATER CORE

NICHROME V
CORE HEATER

SECTION A-A

CADMIUM SHIELD
HELIUM PURGE LINE

OIL IN THROUGH SHAFT ORNL- LR - DWG 5969

HELIUM EXHAUST !
\
OlL OVERFLOW
OlL INLET FOR | }

RING COOLING

SPARK PLUG PROBE

   
   

CENTRIFUGAL PUMP MODEL LFA

- CONTROL GAS
KOVAR SEAL

FITTINGS — PUMP ENCLOSURE

DRAIN TUBE SEAL

FREEZE TUBE
EXPANSION BELLOWS

HELIUM ATMOSPHERE

 

WATER OQUTLET
EXPANSION BELLOWS .

 

 

 

  
 

 

 

 

 

 

 

 

 

 

1 NOSE PIECE - A-J
BORON SHIELD
NOSE COVER

20 in. REF : —

 

 

 

 

 

  
 

JACKET ASSEMBLY HEAT EXCHANGER LINER SEAL
Y5-in. DIA SUPPORT ROD  AND FLANGE
F TURBI CALRODS .
S JE- TUBING FUEL. TUBING \ - 0" RING
a-in. OIA SUPPORT ROD- Ut 7y
*: '":-j”—"——:i"”:—:—:———*a:::"::f;—l'fifii’f AIR INLET
= R ) o
fhkL i T *M/ 4J!
W Y.
G ! wd -
A e ,JV
e P e e aenagd /
" - : e O Al |
ANNULUS FOR g i | /

"KOWLR SEAL

WATER FLOW TRANSITION . Y DIA FUEL TUBING (IN} - J
PLATE SUPPORT ROD FUEL LEADS \PRESSURE FITTINGS
T¢ DIAPHRAGM™ TRANSMITTER CELLS WATER INLET
“HELIUM LEADS TO
DIAPHRAGM

'

- {21t 5in REF W - . e -

Fig. 1. Fluoride Fuel Loop.
 

turbulent flow in the irradiated section in order to
remove the fission heat from the fuel adjacent to
the tube walls (where the velocity is low in laminar
flow).  Tubing having an inside diameter of
0.225 in, was selected for the section in the
neutron flux. In a tube of this size, the fuel mixture
(density, 3.9 g/cc; viscosity, 10 centipoises)3:6
cannot be maintained at turbulent flow (R~ 5000)
through a length much greater than 4 ft without
exceeding the output characteristics of the pump
described above. Accordingly, the irradiated
section was made up of about 3]/2 ft of 0.225 in.-ID
Inconel tubing bent into a U-shape 2 in. across,
and, to keep the pressure-drop low, the connecting
tubing used between the pump and the irradiated

section was 0.40 in. 1D and 0.50 in. OD.

Safety considerations strongly influenced the
design of control equipment and instrumentation
as well as the loop structure, A major consider-
ation was the problem of dissipating fission heat if
fuel should become stagnant in the neutron flux
region. As a precaution, temperatures at many
points were recorded, with provision for shutting
down the reactor in case of excessive values,
A further safeguard was in the fabrication of the
irradiated U-bend from thick-walled tubing for
added heat capacity as well as protection against
corrosion, The thermal capacity thus incorporated
into the irradiated section served to retard the
temperature rise in case of flow stoppage and to
allow time for the temperature monitors and reactor
controls to operate, Inconel fins brazed to the
U-bend assembly and the omission of insulation
from this part of the loop were additional means
of controlling a possible runaway condition,

Personnel safety required shielding around the
pump and the fuel tubing outside the reactor shield.
It was necessary, therefore, to keep the radioactive
parts as compact as possible,

Experience with the First Loop

The first loop experiment was assembled and
was inserted in the reactor after being tested as
thoroughly as possible without actually being
operated with molten salt, The appearance of a
leak in this loop as it was being filled with fuel
demonstrated the need for an operating test prior

 

6S. 1. Cohen and T. N. Jones, Preliminary Measure-
ments of the Densily and Viscosity of NaF-ZrF4-UF;
(62.5-12,5-25,0 Mole %), ORNL CF-53-12-179 (Dec.
22, 1953).

to the rather lengthy reactor installation, The
performance of a preliminary operating test re-
quired that the addition of a drain and fill
connection be made in the final loop design, since
it was not feasible to remelt the fuel after the
test because of the danger of bursting the tubing
from thermal expansion of the fuel (~20% from the
mp to 1500°F). The drain and fill connection is
described below,

The leak occurred in a thin section of a flange-
weld joint after the joint had been satisfactorily
Although this type
of weld had proved satisfactory on other loops,
the combination of stress concentrations at the
weld and of incomplete weld penetration caused
failure in this instance, To provide butt joints of
greater strength, fillet-weld joints were employed
in the assembling of the final loop (Fig.2). Tech-
niques were developed by the Metallurgy Division
for making sound welds of this type on small
tubing without obstructing the bore.

vacuum-tested at temperature,

UNCLASSIFIED
ORNL-LR-DWG t4444

 

 
  

~ %
_______
- -

”
3 5
”
& .
&
>
o
i
P

CROSS SECTION OF FLANGE-WELD BUTT JOINT BEFORE WELDING

     
    
   

b

 

   

Illllllll.‘g DAL
= 5

DETAIL OF
LSS S
s

FLANGE-WELD BUTT JOINT
AFTER WELDING

 

CROSS SECTION OF FILLET-WELD BUTT JOINT

Fig. 2, Welded Joints for Tubing of Fuel Loop.

 
The Basic Loop Apparatus

Beginning at the in-pile portion of the loop, the
fuel system included, first, the irradiated section,
which was essentially a U-bend of tubing, This
component, because of its shape and location at
the in-pile end of the device, was termed the
“nosepiece.’”” From the irradiated section, the
fuel passed through connecting tubing to the heat
exchanger. The heat exchanger was parallel-flow
air~cooled and had a single, central fuel tube with
Between the heat exchanger and the
pump was a simple T-connection which served as
the drain and fill connection. A short section of
the fuel tube between the T and the pump could
be cooled to form a freeze point and thus permit
the fuel to be discharged from the system. From
the pump discharge, the fuel passed through o
venturi flowmeter and through connecting tubing

spiral fins,

from the flowmeter to the nosepiece, completing
the fuel circuit (Fig. 3).

Sofety

Since this was the first experiment of its type
and since the potential hazards were great, safety
considerations prompted many of the design
features. A leak of irradiated fuel would release
radioactive fumes and gases, so, for the protection
of personnel, the loop itself (including the pump
in a hermetically sealed
stainless steel jacket. To reduce the probability
of a leak forming a stagnant pool at the high-flux
end of this jacket, an inner jacket with a gently
tapered end enclosed the nosepiece, This tupered
end formed a sloping bottom for conducting leaking
fuel back from the high-flux zone.

The sealed filled and was con-
tinuously purged with helium to provide an inert

bowl) was enclosed

jacket was

atmosphere around the locop and a means for
detecting fuel leaks by radiation monitoring of the
exhaust helium. The portion of the sealed jacket
inside the reactor shield was double-walled with
channels for circulation of cooling water, This
water jacket protected the reactor structure from
the high temperature of the loop.

The use of temperature as a trouble indication
in the nosepiece has been noted earlier. Since
temperature is one of the controlling parameters of
the corrosion process under study, the importance
of temperature control is obvious, both as a means
of preventing dangerous corrosion rates and to
The temper.
atures at many peoints on the loop, at significant

assure validity of the experiment,

 

points on the water jacket, and in the pump cooling
oil were recorded on several recording potenti-
ometers, At excessive temperatures these re-
corders gave both audible and visual alarms,
The manner in which an over-temperature condition
produced alarms and remedial action by the reactor
and heater controls is described in the section
entitled *‘Instrumentation and Control.”’

The possibility of the fuel mixture assuming a
critical configuration was considered, since the
total charge contained almost 2]/2 kg of U235, 1t
was concluded that criticality did not present a
serious problem, because dispersion of the fuel
through a hydrogenous moderater such as water
would be required.”
both the fuel circuit and the water jacket would
To forestall the
creation of a leak in the water jacket by the fission
heating of a pool of fuel from a possible fuel leak
in the high-flux zone, a network of thermocouples
was focated in the inner jacket around the nose-

Even coincident leaks in

not produce a critical assembly.

piece, The thermocouple indicators were adjusted
to produce an emergency shutdown of the reactor if
the temperature increased 100°F over operating

levels.

A fuel leak which would deposit a large fraction
of the charge near the reactor core would require
excess capacity in the reactor controls to shut the
reactor down, but it was determined that the
contribution to the existing reactivity of the reactor
core would not be large enough to override the
existing control rods.8 [t was concluded that the
possibility of the criticality properties constituting
a hazard depended on an extremely remote coinci-

dence of circumstances.

The presence of radicactive fission products,
some of them gaseous, in the fuel necessitated
the radiation monitoring of the effluents from all
the gas systems in contact with the loop. In
addition to the prevention of dispersion of radio-
active materials through the atmosphere, this
monitoring would give an indication of leaks into
the gas systems. The arrangement of the radiation
instruments for measuring the activity of the
various gas systems is described in the section,

““Instrumentation and Control."” Air in the vicinity

 

’p. Callihan, Nuclear Safety of High Temperature
Test Loop, ORNL CF-53-10-46 {Oct. 7, 1953).

8 etter from M. C. Edlund to O. Sisman, dated March
19, 1954,
 

 

 

 

 

    
  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
   

     
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

» . . ¥
ORNL-LR-OWG 14442
PLANT AIR DEMINERALIZED
WATER
REACTOR OFF-GAS
HEADER —{><}— VALVE
[
O] PLANT WATER —@ REGULATOR
EMERGENCY
RADIATION
NITROGEN
INDICATOR
[
- % PLANT WATER
—~ >4
SAFETY
CHARCOAL RELIEF
ADSORBER PLANT
WATER PURIFIED
HELIUM
ATMOSPHERE Y r@—‘—“ R S /
S / . // S .
| . ,/// / A
| . - ACTOR SHIEI_D s I
AtR . T L . /, S ; /’/”’,/I’ i // ;;1;‘ i
MOTOR |} FREEZE ' s R a1
' HELIUM Al e — — — !
RADIATION > I
INDICATOR P FILL AND ] | il
X e ‘
olL %_g%g Il
RESERVOIR PLANT . TN
WATER [_] %Z T T T T
) o s S S // s
S\
/_/ 4-/ b . ’4 ‘ // . - /,//, / / ), ’ /g /’
J MOTOR|MOTOR // RSN // // s
s - ! Vs ’

 

 

PLANT WATER

 

 

S
DRAIN

HELIUM SUPPLY

 

 

Fig. 3. Schematic Diagram of Fluoride Fue! Loop System.

% . P . -
L // . p

e . <
- S

s

o ,"
S

. s, ,'4/4 s
P // S //4///

   

PR

 

REACTOR CORE/

 
 

of the loop was monitored by a portable, self.
contained air monitor with its own visuval and
audible alarm.

The Auxiliary Systems

The relationship of the auxiliary systems to the
basic fuel circuit is shown schematically in Fig. 3.
The fuel system is outlined in heavy lines, and
surrounding it is the helium-atmosphere jacket,
There were two other helium systems: the flow-
meter operating gas, supplied from the same
cylinders that supplied the jacket, and the pump-
sump pressurizing gas, supplied from high-purity
cylinders, The flowmeter operating gas was
supplied at a
upon the demands of the
This system also discharged to
stack serving the ORNL

The pump-sump gas was oper-

through motor-driven regulators
pressure dependent
pressure cells,
the reactor off-gas
Graphite Reactor,
ated at about 5 psig in order to prevent in-leakage
at the pump and to provide an inert atmosphere
over the fuel in the pump., Since’this helium was
in contact with fuel, removal of fission products
by passage through a charcoal adsorber was

necessary before this gas was discharged to the
off-gas stack.

The other gas systems were: cooling air for the
heat exchanger and operating air for the pump-drive
air motor, and both were arranged to operate up
to the plant supply pressure (120 psi) and at high
flows, The pump-motor air, since it was not in
contact with the loop, was discharged through a

muffler into the LITR building.

In addition to the fuel system, the other liquid
systems were those for water cooling and the one
for the pump cooling oil. Demineralized water was
passed through the loop jacket for cooling, but
plant water was used in the other water systems,
namely, the gas-water heat exchangers for removing
heat from gas streams passing to the charcoal
adsorber or to heat-sensitive back-pressure
regulators, An additional water system was the
water-cooled reservoir for the pump cooling oil.
There were two circuits in this system: one
through the impeller shaft and one through the
flange, both operating from the common water-
cooled reservoir,

 

CONSTRUCTION

The Fuel Circuit

The fuel circuit (see Fig. 1) was made up of the
following components:  pump, flowmeter, heat
exchanger, and nosepiece. All were made of
Inconel and were welded together into a closed
loop of Inconel tubing which was 0.50 in, OD and
0.40 in. ID. The long runs of tubing between the
nosepiece, the heat exchanger, and the venturi
flowmeter were spiraled to allow for some ex-
pansion on heating. The fill line was connected
into the loop tubing about 1 ft in front of the
intake to the pump. The small section of tubing
between the fill line and the pump intake was
fitted with a gas-cooled tube, and this section of
the loop was used as a freeze-off valve. This
freeze point enabled the fuel to be blown from the
loop and the pump by pressurizing the pump. |In
this way the fill line was also used as a drain line,

As has been discussed, a very critical part of
the circulating-fuel experiment was the pump. The
smallest, readily available pump for circulating the
fuel mixture at 1500°F was a sump pump (Fig. 4).

The hot part of the pump was restricted to the part
below the large flange, that is, the pump bowl
(or sump), impeller, and some thermal radiation
shields above them. Above the flange was a long
section which was cooled. This section comprised
the bearings and shaft seals in their housing and
the drive shaft extending out the top. The intake
to the pump was at the side of the pump bowl; the
discharge was at the bottom, A detailed design of
the pump is available but will not be discussed
here. Associated components, such as pump
lubrication and cooling system, drive motor, and
controls, were important parts of the equipment and
are discussed in detail in the section entitled
*“The Pump System,”

The part designated as the nosepiece was at the
opposite end of the loop from the pump (the in-pile
end), The U-shaped piece (Fig. 5) was 18 in. long
and was made of a special heavy-wailed tubing
(0.225 in. ID with 0.125-in.-thick walls), The
radiator fins were furnace-brazed to the U-tube with
copper, as shown in section A-A, Fig. 5. Special

 
 

UNCLASSIFIED
ORNL-LR-DWG 14665

CQOLING

7 OIL IN

=

        
   
   
  
    
  
 

SPEED INDICATOR

_— SPARK-PLUG PROBE

GAS SEAL ASSEMBLY

COQOLING CIL QUT

FLANGE
COOLING OIL

CONTROL GAS

-LINER INSERT FOR VOLUME REDUCTION

LINER ASSEMBLY—""

PUMP INLET - VOLUME REDUCING ATTACHMENT

 

—~RING SEAL
IMPELLER—
T——---- BODY ASSEMBLY
0 1 2 3 4 5 6
Fa— L | | L L
PUMP DISCHARGE SCALE IN INCHES

Fig. 4. Centrifugal Sump Pump for Molten Salts,

 
 

10

   
 

X — "RADIATOR" SECTION OF LGOP - PROJECTED
AND ENLARGED VIEW

CALROD HEATERS — 1 P

S

BODY {16 9%, in. LONG) -
Y

=N

7 A

-7

  

  

  
  

CALROD HEATER - _
PROCESS TUBE ~., P
i

Fig. 5. Irradiated Section or Nosepiece of Fuel Loop.

 

PROCESS TUBE -

P

 

DwG 2Mm22

  
  
  
   

SECRET
DWG 231224

—FiIN

“-FIN{LEFT HAND)
 

adapters were made for connecting this heavy-wall
tubing to the 0.5-in. connecting tubing.

The heat exchanger, with portions of the jacket
and fins cut away, is shown in Fig, 6. The fuel
flowed through the straight center tube, which was
the same size as the main tubing of the loop. Air
was admitted ot the left end and was exhausted
through the exit air line at the right of the illus-
tration. The jocket and expansion bellows formed
a hermetic seal which prevented air from leaking
into the helium protective atmosphere around the
loop. The fins were made from nickel disks which
were shaped and welded together to form a double-
lead helix and were then furnace-brazed to the
tube with gold-nickel brazing alloy. After brazing,
the outside of the finned tube was machined to
permit a close slip fit in the jacket, and the outer
jocket was then welded in place,

Tests of the heat removal capacity of the heat
exchanger were made by an MIT Practice School
group 710 using a steam-to-air system. The heat
exchanger was found to have a capacity of 5 kw
at a 2-psi pressure drop, 10 kw at 5 psi, and 20 kw
at 50 psi.

Fuel flow in the loop was measured by a venturi
of standard design (Fig. 7) which was machined
from a solid rod of Inconel. The unique feature of
the flowmeter was not the venturi but the means by
which the pressures were transmitted to the sensing
device without bringing the radicactive fuel away
from the venturi location, Special pressure
transmitters were designed which were small

 

M. w. Rosenthal, Performance of ANP In<Pile Loop
Heat Exchanger, ORNL CF-54-9-83 (Sept. 10, 1954),

Bp. ). Parry, F. Do Mireldi, snd & D. Rossin,
Momentum uand Heat Transfer Studies n'{ an In-Pile
Lng:f Heat Exchanger, KT-174, EPS-X-211 (Aug. 20,
1954),

EXPANSION BELLOWS

    
    

/f , AR PASSAGE
UBING

f LB | b
'_'12-‘—.--.-144{.1-_4,4:-|
| | | 15
|I|| ’

enough to fit into the avcilable space and yet
could operate at 1500°F. The venturi pressure
transmitter is shown in Fig. 8. A detailed
description of the pressure transmitter system,
including instrumentation, is given in a separate
report.”

All fuel-containing parts of the loop were traced
with heaters, and enough separate circuits were
used so that the temperature of the various parts
of the system could be regulated individually,
For the most part, Calrod heaters were used, and
13 of them were bent to fit the individual com-
ponents, as shown in Fig. 9, There was an
odditional ceramic core heater in the center of the
U of the nosepiece, and the venturi pressure cells
and lines between the pressure cells and the
venturi were heated by specially built units,

The Pump System

The fuel pump has been discussed as part of the
fuel circuit; the auxiliaries to the pump have been
designated as the pump system. The pump was
driven at 2000 to 3000 rpm, through a V-belt
coupling, by a 2-hp Keller air motor. The air
motor was selected because this unit is much
smaller than a variable-speed electric motor, Air
pressure to the motor was controlled monually ot
the panel board. The motor speed was not
measured, but the pump shaft speed was monitored
with a photoelectric tachometer.

Light spindle oil (Gulfspin 60) was circulated by
separate oil pumps through a cooling ring in the
pump flange and through the pump drive shaft and
shaft housing for cooling as well as for lubricating

 

M T. Morgan, Hermetically Sealed High-Temperature
Pressure Transmitter and Hermetically Sealed High-

Temperature Liguid-l.evel Probe, ORNL-1939 (Aug.
29, 1955).
UNCLASSIFIED
SACRET PHOTO 14509

   

NI FINS
EXIT AIR LINE

IR 1L|I||1Lfi-|'w =

TYCOFEYEXEN) ATy I

FUEL TUBE

Fig. 6. Air-cooled Spiral Fin Heat Exchanger,

11

 
     
  

0.4375 —in. DiA

UNCLASSIFIED
ORNL-LR-DWG 15020

i 2
\ . |

INCHES

Fig. 7. Venturi for Fue! Loop.

the pump. The oil circulating system is shown in
Fig. 10. Oil in the reservoir was maintained at
about 35°C with water cooling coils. Flow was
maintained at about 1 gpm in both circuits. Low-
flow alarms were installed at the oil reservoir,
and a standby pump was piped into the system to
ensure oil flow. The oil reservoir was vented to
the reactor off-gas stack after a temporary plug in
the pump off-gas line forced a small amount of
gaseous activity into the oil.

The space above the fuel in the pump was
continuously purged with pure helium gas; the
off-gas from the pump will be discussed as part
of the loop off-gas system,

The Off-Gas System

The sources of gas which could carry radio-
activity in the event of an accident were as
heat-exchanger cooling air, loop-jacket
purge helium, external-lead-shield exhaust air,

Also,

fission gases are carried off by the pump purge

follows:

and the flowmeter pressure-cell helium,

12

helium, All these gases were handled through an
off-gas system which is diagramed in Fig. 3.

The heat-exchanger cocling air was vented to
the reactor off-gas stack. This air could manually
be diverted through the charcoal trap, but it was
not intended that large volumes of hot gas ever be
put through this adsorber. Flowmeter pressure-cell
helium and the lead-shield exhaust air were also
vented to the stack, and no provisions were made
todivert these gases through the charcoal adsorber,

Helium was bled into the loop jacket at the
nose of the loop .and was taken off at the pump.
Gas at a rate of about 2 ft3/hr and at 2 to 4 psig
was used to purge the loop jacket. The exit gas
was passed by a radiation detector before being
bled to the stack. If this gas had become radio-
active it would have been manually diverted to the
charcoal adsorber, The jacket was also equipped
with a pressure relief valve set to release at
9 psi and connected through large tubing directly
to the off-gas stack. The relief valve prevented
pressure buildup in the jacket, either from an
El

UNCLASSIFIED

50-A-1109A UNCLASSIFIED
DBNL-LR-DWG 50654 PHOTO 14140

HELIUM GAS LINE MODIFIED CHAMPION V-4 SPARK PLUG

TO CONTROL PANEL 7 WITH PLATINUM TIP

WASHER

T

%
— b4 PLATINUM CONTACT

e

S

       
 

 

 

   
     

    
 

         
  
 

 

 

 

 
 

 

     
  
  

e i'rl )
R R S .
« e S N = —
>
T A

 

LIGUID LINE
T VENTURI

Fig. 8. Pressure Tronsmitter for Fuel Loop.

 

 
 

vl

UNCLASSIFIED
ORNL-LR-DWG 15058

 

T.C's NOT SHOWN ON DIAGHAM

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TC. PANEL ON MAIN CONTROL BOARD ] HEATER NO.| RATED POWER | VARIAC NO. | CONNECTOR NO.
RECORDER TRA-5 " [ PaT cont T 00 2a00w__ | 17 CN-7
REC. PT.  TC NO. LOCATION RECORDER TRA-1 | TRA-2 TRA-3 1= 1500w
1 80 LINER {END) (OTHER LINER TC's' NOS. 81 AND 82) “RecoRDERPT |1 |2 (31123 11231456 7][6]s —5 5 " recow— -8 CN-8
2 83 DEMIN. WATER TO LINER ~ Tc.Ne. | 2ales|  lei]e2| |e32728|29]30|3t|35]33] 36 —xae3 T T Thoow ] e | cNg |
3 84 DEMIN. WATER FROM LINER HECORDER TRA -3 — TRA4 ] e 2soow | Tag | v ]
4 85 OIL RESERVOIR RECORDER PT. qoi wizl1]2|3][a]s]e[7]8]9]t0lu M Eals 2500W T2 cntz
5 86 OIL COOLING RING (ON OIL LINE) TC.NG. | 34)139] 40| 4314647/ 48[49]50(51 [56]59/63 (6568 T en-6 Toooow | T4 | cn-4
6 87  PUMP SHAFT OIL (ON OIL LINE] RECORDER TRA_S SPARES | 2T rs0w i NS
7 88 AR INTO HEAT EXCHANGER RECORDER PT. |1 | 2| 3|4i5|6[7] B9 [10][n 12 ] T ogalE 70w 1 Te CN-6
8 89 CHARCOAL ABSORBER TC.NO.  [80783[84 ssleeleflsa\se 90| 91 [92]9328] | 5.9 2000W T cN3
9 90 COOLING WATER FROM PUMP SHIELD —— AR | eoow | o
10 91 PROCESS WATER INLET TeNo  |37lal [42|44|45;ss‘60[52!53]54{55\57[58|61 [e2 AT wooow | T-1 ot ]
11 92 EXHAUST AIR FROM PUMP SHIELD SPARES _IFZA—iE flNTSOW oy - ch-zZ -
12 3 QUTSIRE OF PUMP SHIELD | Tcno 54 166 ‘6?? . l 1‘ l [ I i ‘ [ I I ‘ 3a-13 SBOw T3 CN-13
*ALS0 TC's B AND 82; ALL THREE ON A SELECTOR SWITCH. 14814 - T-14 CN-14
. 1525 2500w T-15
SYMBOLS: {616 2500W T-16

 

 

 

 

 

 

@ THERMOCOUPLE

HEATER

DIRECTION OF FLOW
THERMOCOUPLE LOCATION (THERMOCOUPLES ON HEATERS MARKED "H")
THERMOCOUPLE WELL LOCATION (ALSO MARKED BY "w"}

.,,T@

v ?7

 

 

 

% Al ¥ [
%7 WW e v VV Y W

DISTANCE FROM LINER SEAL AND FLANGE (FEET)
6 5 4 3 2

Tl e i = e Py mmm g e - - s o

0 $ 2 3 4 5 6 7 10 1 1?2 13 14 15 TS
DISTANCE FROM INNER END OF LOOP (FEET)

 

 

 

 

 

Fig. 9. Location of Heaters and Thermocouples on the Loop.
 

UNCLASSIFIED
ORNL-LR-DWG 14443

PUMP SHAFT

 

 

SHAFT OVERFLOW

COOLING
RING

 

 

 

 

VALVE KEY

NORMALLY OPEN

THROTTLING

 

 

N bl

 

 

]

ROTAMETER
LOW-FLOW ALARM | OVERFLOW

I i X

NORMALLY CLOSED

 

 

 

 

 

v| RroTaMETER
LOW~FLOW ALARM
Yo
/

 

 

 

 

 

 

ALARM
COOLING
WATER -
&} - ( vL
Py '
=) OIL RESERVOIR
=) | TEMPERATURE ALARM L
| < M
| I
DRAIN

 

 

 

 

sl Joo 1 Jov I

 

Fig. 10. Oil Circulating System for Fuel Pump.

obstructed line or from a water leak into the high-
temperature region,

High-purity helium was continuously bled across
the surface of the fuel in the pump bowl at about
2 ft3/hr and at 3 to 5 psig to provide an inert
atmosphere and to remove gaseous fission
products. The activity of this gas was too high
for the gas to be vented directly to the stack, so
it first passed through the charcoal adsorber,
which retained the fission gases long enough to
permit most of the activity to decay before it was
released to the stack. The charcoal adsorber
(see Fig. 11) was a 6-in. pipe 10 ft long, con-

taining about 2 ft3 of 12- to 14-mesh Columbia
activated charcoal, The adsorber was placed
vertically in a hole in the reactor floor and was
shielded at floor level with lead and concrete,
Gas entered at the bottom of the adsorber, so the
most radicactive part was deep underground, The
adsorber operated at the temperature of the sur-
roundings,

The Assembly Procedure

When completely assembled, the ‘‘in-pile’’
portion of the loop was enclosed in a stainless
steel jacket, as shown in Fig. 12. The flange

15
UNCLASSIFIED
CANL-LA-DWG 15314

B=in. DIA
STAINLESS STEEL PIPE

THERMOCOUPLE WELL —— |f

|
STAINLESS STEEL #---J: f
SCREEN =

Figs 11. Charcoal Adsorber for Loop Off-Gas System.

   

which closed this jocket was used as the bench
mark from which the position of all components of
the loop was established. The first step in the
assembly of the loop was to weld together all the
fuel system up to the flange, This was done by
assembling the in-pile components onto the flange,
or in the proper relation to the flange (with the oid
of the support rods which were attached to the
flange), and then making the welds in the proper
sequence (see Fig. 13).

All welding was done under close supervision
and inspection. As each weld was finished it was
inspected for cracks by the dye-penetrant method
and for internal construction defects by x ray,
Each section was also checked with a helium leak
detector as it was assembled, To allow for thermal
expansion of the loop, the only point of fixed
attachment of the loop was at the pump, The
heat-exchanger air inlet and outlet tubes were
attached at the jocket flange with the use of
bellows, and the rest of the leads were flexible,

Next, the heaters were installed on the in-pile
portion of the loop, and then the thermocouples
were installed. Calrod heaters were bent to the
proper shape and were attached to the loop tubing
with hose clamps. To ensure that there were no
gaps or overlaps in the active sections of the
heaters when they were attached end to end, the
heaters were x rayed to determine the actual length
of the heater element, The terminals of the heaters
were modified to permit operation at higher temper-
atures by replacing the existing terminals and mica
washers with welded connectors and washers made

FHOTQ 129980

 

 

Fig. 12. Assembled Fuel Loop in Water Jacket.

 

 

Fig. 13, Bare Fuel Loop with Support Framework.

16

 

 
 

of Lovite, The welding connectors are shown in
Fig. 14. The electrical leads were insulated with
a double layer of woven glass sleeving (and with
ceramic beads in very hot sections).

Thermocouples were made from Chromel-Alumel
duplex, nonimpregnaoted glass-insulated wire. |In
regions where the thermocouple leads
subjected to temperatures above 1000°F, the glass
insulation was removed and replaced with ceramic
beads. The thermocouples were formed with a
capacitor discharge welder by flashing the ends
of the wires against a block of graphite, The
quality and ductility of the bead was greatly
improved by shielding the weld with helium during
flashing. The thermocouples were attached to the
loop tubing ot the desired points by using o similar
technique, The thermocouple was held with
tweezers, which served as one electrode, and then
pressed against the loop tubing, which formed the
other electrode.

were

UNCLASSIFIED
ORNL-LR-DWG 16682

 

Fig. 14. Welding Connectors for Electrical Leads.

After attaching heaters and thermocouples, the
first layer of insulation, quartz-fiber tape, was
applied, Fiber-frax insulation (fibrous aluminum
silicate bats, The Carborundum Company) was
pocked around this first layer, and, finally, glass-
cloth tape was used to bind down the Fiber-frax.
Mo insulation was applied to the nosepiece section
of the loop.

Shaped graphite blocks were fitted into the
vacant spaces around the insulated loop from the
nosepiece section back to the heat exchanger,
The graphite served os a shield to attenuate the
radiotion from the reactor during installation of
the loop. Graphite was chosen for this shield,
since it was easily cut and would withstand high
temperatures., It was found necessary to coat the
grophite blocks with sodium silicate in order to
prevent the graphite from rubbing off on the
electrical leads and thereby lowering the insu-
lation of the leads to ground,

To prevent fission heating in the larger-diameter
tubing, ferro-boron and cadmium foil were used
to provide thermal-neutron shielding for this
region. The outside of the water-cooled jacket
was wropped with cadmium foil from the step
in the jocket to within 2 ft of the in-pile end.
The ferro-boron was applied in the form of disks,
interspersed with graphite., Figure 15 shows the
insulated loop with some of the graphite installed.

The jocket, which provided water cooling and
an inert atmosphere for the loop, was fabricated
from Y _in.<thick stainless steel (see Fig. 12).
Two baffles, running the full length of the annular
region between the double walls of the jacket,
directed the water flow so that the cooling water
went in along the bottom half of the jacket, across
the end, and returned along the upper half, The
flange, which is part of the loop assembly, was
bolted onto the jocket during final assembly, A
hermetic seal was made with a copper gasket,
Electrical and gas-line connections to the in-pile
region of the loop were made through the flange,
using Kovar-glass seals for the electrical and

PHOTO 12998C

 

Fig. 15, Fuel Loop with Heaters, Thermocouples, Insulation, and Part of Shielding.

17

 
thermocouple leads, The loop tubing and heaters
were carried through the center of the flange,
encased in a thin-walled ]]fz-in.-diu tube which
connected the water-cocled jacket to the pump
enclosure, The pump enclosure was not water-
cooled but served to contain the helium protective
aotmosphere and to act as a safety receptacle in

case of a fuel leak,

The complete assembly was long and awkward
to handle but was even more awkward when the
pump was installed, because the pump was heavy
and needed external support, The in-pile portion
of the loop was therefore completely finished and
was encased in the jacket, and all electric and
thermocouple lead connections were made at the
Kovar seals and tested (the electrical circuits at
500 v) before welding the pump into the loop.
First, the pump bowl was welded into place, and
the fill and drain line was attached; then the
heaters, thermocouples, and insulation
installed as described for the internal portion of
the loop (see Fig. 15). After final inspection and
testing, the pump enclosure was welded into place,

were

 
  
  

 

 

 

Instrumentation and Control

The main instrument panel is shown in Fig. 16.
Heater power controls, temperature recorders and
controls, air and helium flow control, jocket-
cooling-water flow control, fuel flow control,
radiation alarms, and safety circuits were located
on this panel, A schematic instrument flowsheet
is shown in Fig. 17.

The heater layout has been discussed as part of
the fuel circuit; heater locations are shown in
Fig. 9. The Variac controls for the heater circuits
were mounted on the panel board, Only manual
control was used for power to the heaters, except
for the nosepiece heaters, which were controlled
by a proportioning-type temperature controller,

The purpose of the voltmeter and the ammeter
at each Variac was to indicate the power being
supplied to each heater circuit, It was found that
the type of ammeters used did not read correctly
because of their close proximity to the Variacs.
Attempts to shield the meters from the magnetic
field of the Variacs were only partially suc-
cessful, so the panel-board ammeters were used

only as an indication of current flow. Jacks and

   

UNCLASSIFIED
PHOTO 12957

i e, P TA

 

 

".m—-‘ .‘_ h -
esips . L =

  

oo oo
=

 

  

 

 

 

 

 

 

 

 

 

Fig. 16. Instrument Panel for Fuel Loop. (Secret with caption)

18

 

 
 

 

 

 

 

i

PR o

®

|

HEL!IUM { COMMERCIAL) CYLINDERS

2

A

— JACKET (WATER FILLED)

 

 

\
i
|
!
!
|
I
I
|

LOOP - (IN PILE HOLE HB2)

i
4

INLET TO HELIUM PURGE
FOR ANNULUS OVER LOOP
PUMP SYSTEM

 

 

HELIUM (PURE} CYLINDERS

 

 

 

 

 

VENT

 

Vaaq v

 

 

 
   

   

 

—
DOWNSTREAM UPSTREAM

(LOW) (HIGH) ~_
N

—iF t
FE-3 '
PUMP .
TO VACUUM  y 5y vas
PUMP
VENT LINE_ 7

AR MOTOR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
 

 

    

 

 

 

 

 

 

 

 

 

 

L EXHAUST TO
- 1 psi | OUTSIDE
COMPARTMENT PRESSURE HELIUM QUTLET V54 Vi
NNECTED TO PUMP SHELL - g«
co -t
PLUGGED
TO COMPARTMENT ugs V54 1 (cooLer
HEAT EXCHANGER NO.1 1 FILL TUBE
; \ - AAANAAARP -  _
t = VVEENVVVY V22
O o % 3 |HEAT ExCHANGER NO. 4
. : . , : - — /P | OiL STORAGE
' I | ! n | \ 3/ TANK = IONIZATION
f | I t I t P CHAMBER
(TIAY /TIR (TIAY (TIR (TIAY Lo
‘ g/ % SEPARATOR
Lo L] V55
/XA T
=
N &
vi4 HEAT EXCHANGER NO.3 <5,
SETTING I
9.3 psi i “Is
PLANT [ “I> ——
WATER -
v Z8 rfl
v 15;
PLANT VENT LINE
WATER
Y
@ I
9 .é V23 | VY
AANAAA '.fl
l VVVVVVY i VENT LINE
' HEAT EXCHANGER NO. 2 ‘
v a7 ves 30 psi @
X WATER V30
25 psi
AIR SUPPLY
20 psi
DEMINERALIZED WATER

 

 

 

 

 

 

 

 

 

 

 

Fig. 17. Flow Diagram for Instruments and Controls of Fuel

TO STACK

 

 

Loop. (Secret with caption)

UNCLASSIFIED
ORNL-LR-DWG (5458

LEGEND
PROCESS PIPING
INSTRUMENT PROCESS PIPING
ELECTRICAL LINES
INSTRUMENT AIR LINES

PRESSURE REDUCER OR BACK PRESSURE REGULATOR
(DEPENDENT ON CONNECTIONS) WITH VENTED EXHAUST.

OTHER SYMBOLS CONFORMING TO 1.S.A. STGNDARDS (PUBL. RP 5.1)

 

   
   
  
  
   

LITR
OFF -GAS
LINE

 

1B50 psi

 

NITROGEN CYLINDERS

PLANT
AlR SUPPLY
100 psi

INSTRUMENT AIR
SUPPLY

19
switches were installed so that accurate current
readings could be taken periodically by plugging
in a precision ammeter at each circuit,

It was desirable to have in use a large number
of thermocouples during heatup of the loop and
enough spares so that loss of a few thermocouples
by burnout would not hamper the operation of the
loop. Thermocouple locations are shown on Fig. 9.
Continuous recordings of the temperature at 36
points on the loop were made by three 12-point
recorders (one not shown on the instrument panel).
A fourth 12-point recorder was used to keep a
record of low temperatures, such as those for the
coolant water, coolant air, pump oil, lead shield,
and charcoal adsorber, In addition, the maximum
nose temperature was continuously recorded on
two separate one for the pro-
portioning-type controller and the other for an
over-temperature safety circuit, Each of these
single-pcint instruments was provided with spare
thermocouples that could be switched in if
desired.

instruments -~

Pump-motor air and heat-exchanger air were
both controlled manually at the panel board, The
air pressure to the pump motor was adjusted to
give the desired reading on the venturi flowmeter,
and the pump shaft speed was also indicated by
means of a tachometer. Reduction in flow with
no reduction in tachometer speed would indicate
either an increase in fuel viscosity or plugging of
the loop. Since failure of the plant air supply
would be disasterous to the experiment, a standby
supply of nitrogen (six cylinders —~ 15- to 30-min
supply) was set to cut in when the plant air supply
dropped below 75 psi.

Cooling water to the loop jacket was monitored
at the panel board, but flow adjustment was made
at the supply, since adjustment was not necessary
after the experiment was started. Demineralized
water was used, but plant filtered water was also
tied into the system, to be used if the supply of
demineralized water was cut off.

Fuel flow was continuously recorded by means of
a Speedomax. The venturi flowmeter has been
discussed as part of the fuel circuit. Helium gas
supply to the flowmeter pressure cells as well as
pump purge helium and loop-jacket purge helium
were controlled at the panel board.

In addition to an air monitor behind the loop
equipment, the controls for two radiation instru-
ments were mounted in the panel board. One was
connected to an ionization chamber at the line
which carried all exhaust gases to the stack, and

20

the other monitored the activity in the jacket purge
helium and in the heat-exchanger exhaust air,
Since most of the gas was vented to the off-gas
stack manifold, which was at a negative pressure,
it was desirable to use back-gressure regulators on
those lines where the flow was controlled. The
heat exchangers shown in the instrument flow
diagram (Fig. 17) were used to cool the hot gases
before they passed through the back-pressure
regulators, where they might damage the rubber
diaphragms,

As has been discussed, one of the primary
considerations in the design of this experiment
was the safety of personnel and the reactor. The
heart of the safety system was the annunciator
panel, which was part of the instrument board,
The safety circuitry is shown schematically in
Fig. 18. Certain conditions were recognized as
being unsafe for either personnel or the experiment,
For each of these conditions there was a light on
the annunciator panel. When everything was
operating properly, all these lights were lit, as
was a green master light. When a condition
became abnormal — for example, a nose temper-
ature over 1525°F — the nose-temperature and
master lights went out, an alarm light went on,
and a buzzer was activated, The buzzer could be
cleared for a predetermined set time but not
permanently unless the trouble was corrected, If
trouble became worse — say the temperature con-
tinued to rise — a second light went on, the
reactor was set back, and the loop heaters were
cut off.
give an emergency shutdown (scram) of the reactor.

Further trouble in certain cases would

For each condition of trouble the action taken
depended on the seriousness of the trouble.
Potential sources of trouble and the action taken
are given in Table 2. Additional alarms which
were located at the site of the detecting element
rather than on the panel board are listed below:

Low pressure of nitrogen cylinders
Low air flow te pump motor

Low oil flow to pump shaft

Low cil flow te pump ring

Low pressure of helium cylinders

Thermocouple failure on recorder indicating the

temperature of in-pile end of loop

These were buzzer-type alarms that called the
operator's attention to a condition which needed
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED
ORNL-LR-DWG (5318

 

 

 

 

 

 

 

 

 

 

 

 

 
   

 

 

  

  

   

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 18. Wiring Diagram of Safety Circuits of Loop Controls.

34
HOT QD ’ S »
%KIB% Iso‘% 62% 64‘% :[ 68 l
KI6-1 TRA-{ | TRA-2 _{TRA-3 XA-1 RECESSED o o L p—
T AMBER Kig-4 Sw3 w3 Sw3 K3-2 = SCRAM == K20-1 #EKI9-3 = K2A-2 SW7-1{TPST)
Ki7—1 52  PB. SAFETY CIRCUIT
K19 -1 I & 82 85 &% HKi-2 BY-PASS SWITCH
. : FOR WARM -UP
§7-2 PERICD
NC.
7
u.c XA-1 SW2 "
8 TROUBLE ¢ Bo 4|j)_ 43 HK20-2
TRA-1 SW2 ACKNOWLEDGE _|—  ==ka-1 s¢Kk4-2 PXa-4 PXA-1
9 PB. *s3 Swa ESM Swa
> TRA-2 SW2 46 42 44
& iO} ~
> TRA-3 SW2 o
* 11 g
oW TRA-4 SW2 w
<3 12 S @ SCRAM st sz 55
> g E— RESET
2 e 13 » P8
=z FXA-4 SW2 = V-8 V-5
L “ SOLENGID SSOLENOID Rt
E FXA-3 SW2 - BUZZER (AIR) (Na)
a 15 = K17
5 16 S . USUALLY 52 54 56
ENERGIZ
UG, = RA-( SW2 o ED
EXPERIMENT |7§ TROUBLE
power (1) UC = RA-2 SW2 TROUBLE
RED 18 ACK 39
-] LA-1
Ke-3 FAST SETBACK swa
RESET P.B, 40
K16 K18 K19 K20
AMBER K2 K2A K3 K3A USUALLY USUALLY
34 USUALLY ENERGIZED ENERGIZED ENERGIZED
L
NEUTRAL £AST SETBACK SCRAM LOCAL ALARM HIGH TEMP  LOWLEVEL HIGH LINER LOW PRESS. AR EMERGENCY HEATER
WATER COOLED SUMP FUEL  PRESS.  AIR SUPPLY SUPPLY N, SUPPLY CONTROL
56 JACKET CONTROL CONTROL
ALARM —- AUXILIARY RELAYS e
RESET PB.
lg,é LZE%—J lzgé lmé lssh
TRA-3 TRA-5 | TRA-4 FRA- FXA-2 RA-1 | RA-2 XA~
K7- K8-2 - L. - —
e SW1 SW1 Swi Swi KI0-2 Swi F-KIS-2 | K14-2 o= oy T swi Swi
O 32 i 345—
NOTES
1 ALL RELAY CONTACTS SHOW WITH RELAYS DE-ENERGIZED.
HIGH TEMP HIGK TEMP HIGH TEMP HIGH TEMP LOW FLOW LOW FLOW LOW PRESS. HIGH RAD. LEVEL FUEL LEAK 2 S4 AND 55 OPEN MOMENTARILY WHEN TIMER TIMES QUT.
CONTROLLER LOOP END HOT SECTIONS OF tOOP COLD SECTIONS OF LOQP FUEL JACKET WATER AIR SUPPLY VENT SYSTEMS XA-1
TRA- TRA-2 TRA-3 TRA-4 FRA-1 FXA-2 PXA-1 RA-1,-2 HIGH LINER PRESS. 3 SPARE TERMINAL CONTACT NOS. - 19, 20, 47, 48, 57, 58,
WATER-COOLED JACKET PXA-4
TRA-5 59,70, 71, 72, 73, 74, 75, 76,77, 78.
_ HOT ~
l J_ J_ —ot BILL OF MATERIAL 4  ALL RELAYS USUALLY DE-ENERGIZED DURING NORMAL
K2-1 K3-3 Ki-1 K2-2 3¢ SYMBOiL?JEI‘TT.I“F]CjA;I"ICVJVN— . oo DESCRIFTION OPERATION UNLESS INDICATED OTHERWISE,
FAST SETBACK : NC.3-
*3 {MAKE-UP} ¢ i ARROW-HART & HEGEMAN, 110V AC COIL, 3PU-N.C.;2-NO} 5 CONTACTS ON ALARMING DEVICES (PRESSURE, FLOW
K5-1 Ki,K2,K2A, K20, K3A, KI6 POTTER BRUMFIELD, TYPE MR-1tA, HO V AC COIL, DPDT AND LEVEL SWITCHES, RECORDER SWITCHES ETC.,
Ke-1 C.R.ANNUN K4 GUARDIAN, SERIES 200, 10-VAC COIL, DPDT EXCEPT FOR RELAYS SEE NOTE 1) SHOWN IN POSITIONS
.y . : 03 K6, K7, K9, KIO, K12, KI3, K5 LEAGH, TYPE 324 ASSUMED DURING NORMAL OPERATION.
K-t K8, Ki1, K14 LEACH, TYPE 324 & SWITCH S7 CONTACTS SHOWN IN UNBLOCKED (OR
SCRAM TO CCNTROL ROCM POWER 51,52, 53, 56,58 P.B. SWITCHES, 2 CIRCUITS {1-N.O.; 2-N.C.)
KS-9 04 AND SAFETY CIRCUITS NORMAL OPERATING} CONDITION.
KO-t LIGHTS 1,2, 3, AND 6 THRU 1B | NE =51, PLG 849
Kit- 1 LIGHTS 4,5 6W, 125 v, S6 CANDEL ABRA BASE LAMP 7 DURING NORMAL OPERATION OF EXPERIMENT
FAST SETBACK TIMER f CYCL-FLEX, HP 2646 SOLENOID V8 AND V5 ARE DE- ENERGIZED, SOLENOID
RED )5y 1 ) ’
CONTROL ROOM @ K2h -t {DROP-0OUT) T i STANCOR PC 134 VALVE VB WILL THEN BE OPEN AND SOLENOID
POWER -
USUALLY BUZZLR LUNGEN, NO. 415 SIXEZ VALVE V5 CLOSED
ENERGIZED @ AMBER K18 { ARROW-HART 8 HEGEMAN QO VAC COIL, 3P(1-DPDT; 2-N.O.)
Kl K17, Ki9 ARROW- HART & HEGEMAN, 10 VAC COIL , 3P(1-DPDT; 1-N.0.: 1-N.C.) 8 SYMBOL DEFINITION:
NEUTRAL | . : ! e U.0. MEANS USUALLY OPEN.
e —C6/ 57 | swiTcH, TPST U.C. MEANS USUALLY CLOSED.

21
 

Table 2.

 

Safety Features of the Control System

 

Protective Action

 

 

Condition Cutoff of Cutoff of Heaters, Heat
Alarm  Setback  Scram Heaters Exchanger, and Pump
Low water flow through liner jacket x x x
High temperature at liner jacket x x x
Low air flow through heat exchanger x x x
Low pressure of air supply to pump motor x x x
and heat exchanger
Radiation from heat-exchanger exhaust x x x
High temperature of in-pile end of loop x x x x
High temperature of other locations on loop x x x x
Radiation from helium vent and off-gas lines x x x
High pressure inside liner X x x
Leak indication by thermocouple grid detector x x x x

 

correcting but which was
hampering the experiment.

not yet seriously

Radiation Shielding

More than half the fuel was outside the reactor
shield at all times, and the fuel travel time from
the nosepiece to the pump was only a few seconds
(see Table 1); so it was necessary to shield a
high flux of delayed neutrons as well as a very
intense gamma source. The shield design was
based on calculations which are shown in another
report.'2 It was intended that during operation
the shield would be 15 in., of lead followed by
12 in. of paraffin and a sheet of Boral. It was
found, however, that the quantity of capture
gammas produced in the paraffin was so high that
an additional shield of high-density concrete block
(maximum of 2 ft) was required outside the
paraffin,

The shielding was of two kinds — permanent
lead and demountable shielding. The permanent
lead was that required to shield the experiment
24 hr after reactor shutdown (6 in. of lead). It
was designed to be used during installation and
removal of the experiment as well as during
operation of the loop. Attached to the face of the

 

’2W. E. Brundage, W. W. Parkinson, and O. Sisman,
Questionnaire for LITR Fluoride Fuel Loop Expert-
ment, ORNL CF-53-12-140 (Dec. 17, 1953).

22

reactor was the utilities shield (Fig. 19). This
shield had a door for shielding either the empty
beam hole or the in-pile portion of the loop after
the pump end had been removed. For removal of
the pump portion of the loop, a hydraulic shear
was built into the utilities shield (Fig. 20), At
termination of the experiment the connecting tubes
between the pump and the in-pile portion of the
loop as well as all the heater and thermocouple
lead wires and a few copper tubes were cut at
this point. Provisions were also made in the
utilities shield to bring out water and air lines
and electrical wires through lead plugs.

After installation of the experiment, the pump
shield was fitted into place behind the utilities
shield (see Fig. 19). The remaining lines and
wires were brought out through this shield, but its
primary purpose was to house the pump and to
serve as a support for the pump drive motor, This
shield was also the one in which the pump was
transferred to the hot cells, Both the pump shield
and the terminal shield were cooled by water which
passed through cooling coils cast into the lead.
The pump shield was on wheels but was held in
place with screw jacks.

The first layer of demountable shielding was
8 in. of lead brick, which was installed after all
external connections were made to the loop.
 

 

PUMP SHIELD

o

 

 

UNCLASSIFIED
PHOTO 12780

UTILITIES SHIELD K

- o TTREE

Fig. 19, Pump ond Utilities Shields ot Face of Reactor.

Figure 21 shows the loop during charging of the
fuel. The lead bricks are in place on all three
sides of the permanent shields, When the charging
line was removed and the lead-brick shield was
completed, paraffin blocks were stacked around
and on top of the lead bricks, and then a wall
8 ft high of high-density concrete was built up

When the reactor was started
up it was found necessary to erect a wall of
concrete blocks in front of the main control panel,
but this wall was spaced about 3 ft in front of the
panel to permit access for occasional adjustment
A lead-glass window in the wall
permitted viewing of the instruments,

around the paraffin,

of the controls.

23

 
 

24

 

SLIDING DOOR

 

— J “ UNCL.A;’:FIIED

PHOTO {2547

    
  

i i
{-----n.-——.‘.q..

o

..—l--l-ul.-l-F-— e i i .

HYDRAULIC SHEAR —

-
L

   
 

  

ra

»
&£

UTILITIES SHIELD

e i

e L sl * = 2 - =

Fig. 20, Hydraulic Shear Mounted in Utilities Shield.

 

 

 

essessesTreTERE 0 9 o4

- R

 
 

 

Fig. 21. Demountable ond Permanent Pump Shields.

 

r

i

'-fl

UNCLASSIFIED

PHOTO 12781

 

25
 

OPERATION

Leak Testing and Cleaning

As soon as the fuel circuit was completed by the
welding of the pump bowl to the loop tubing, the
assembled circuit was leak-tested as a unit. The
pump bowl was closed by clamping onto it a plate
having a connection to a vacuum pump and a helium
leak detector., The fuel circuit was bathed with
helium and was proved to be leak-tight. The
interior of the loop was cleaned by circulating
absolute ethanol through the loop by means of a
small laboratory centrifugal pump. The laboratory
pump was connected by plastic tubing so as to
take suction from the pump bowl and discharge
through the loop-pump discharge port, The ethanol
was drained, and heaters, shielding, and insulation
were attached to the loop structure as described
earlier., After the loop was inserted in the water
jacket, the plate and vacuum testing equipment
were replaced on the pump bowl for leak testing at
high temperature. The loop was evacuated and the
water jacket purged with helium to provide an

CHECK VALVE

   

 

 

 

 

inert atmosphere inside and outside the loop,
which was heated to about 1500°F by its own
heaters and tested with the helium leak detector.

In order to test for leaks under operating con-
ditions and to clean up oxide scale inside the
loop, an approximately equimolar mixture of NaF
and ZrF, was circulated in the loop for several
hours prior to installation in the reactor. The
charging equipment is shown in Fig. 22 and con-
sisted of a furnace holding a charge tank connected
to the loop and to a cylinder of high-purity helium,
The regular helium supply for the pump was
supplemented by an oil-filled bubbler on the outlet
and by a sensitive pressure gage on the inlet
during filling.

After the loop and fill line had been heated to
between 1400 and 1500°F, the pump was started,
and pressure was applied to the molten fluoride in
the charge tank, Salt was forced into the loop until
electrical contact was made with the upper probe
of the two probes in the pump. The mixture was

UNCLASSIFIED
SSD-B-1113A
ORNL-LR-DWG 5069A

FREEZE-OFF LINE

 

 

 

 

CHARGE TANK

   
   
  

HIGH-PURITY HELIUM DRAINING
PRESSURE SYSTEM

 

 

MELT-OUT FURNACE ——®=—i

 

 

 

 

   

HIGH-PURITY HELIUM

Fig. 22, System for Filling Loop with Fluoride Mixture. (Secret with caption)

26
 

circulated fo‘i’flrflé hr at about 1500°F and at various
flow rates to correlate pump shaft speed with flow,
Circulation was then stopped, and the fill line to
the charge tank was reheated in preparation for
draining the loop. When the freeze-off section of
the loop was cooled to about 1000°F, pressure
was applied to the pump bowl in order to drain the
loop, which was allowed to cool and was then
prepared for transfer to the LITR,

Installation in Reactor

The loop was inserted in hole HB+2 through the
vtilities shield, and connections to previously
installed piping for helium, air, water, and off-gas
systems were made during a protracted reactor
shutdown, The pump shield was moved into place,
and wires from heaters and thermocouples were
connected to the proper panels prior to the place-
ment of the removable top of the pump shield, The
hydraulic shear was located on the tubes and wires
it was to sever, and the permanent shielding was
closed by insertion of the top plug of the utilities
shield. Completion of assembly of the permanent
shield permitted the mounting of the pump drive
motor and connection of the remaining services to
the experiment. After pressure checking the loop
for leaks, the filling system, with a fresh tank of
the NaF-ZrF, mixture, was connected to the loop
fill line,

Filling

The loop was filled with the salt mixture for a
final operating and leak check after installation,
before committing the enriched uranium charge to
so complex a system, This checking process also
constituted a second cleaning operation., Filling
and draining were carried out as before, and the
loop was operated on this charge for 8]/2 hr.

The charge tank of nonuranium salt was then
replaced with a container of the enriched uranium
fuel mixture, and the loop was filled with fuel.
The material in the filling line was allowed to
freeze, and the line was cut off and capped with a
Swagelok fitting. The fuel container was weighed
before and after filling the loop. The weight
change indicated that 4.98 kg were charged into
the system, with probably 4.95 kg of the fuel in the
circulating system and 0.03 kg frozen in the fill
line,

While the fuel mixture was circulated through the
but before the reactor was started, the
lead bricks were stacked around the

loop,
remaining

former location of the fill line. Paraffin and Boral
were arranged around the lead in the first shield
configuration. It was noted that, in the confines
of the work space at the reactor face, erection of
stacked brick or block shields was quite time-
consuming. An assembly of precast slabs, possibly
of concrete alone, would have saved a large portion
of this time,

Reactor Startup and Operation of the Loop

After the lead and the paraffin were stacked in
the first shield arrangement, the reactor was
brought up to 1% of full power, Radiation measure-
ments indicated that the shield was inadequate
for capture gammas from the delayed neutrons in
the paraffin. The paraffin blocks were then re-
arranged, and concrete blocks were added (Fig. 23).

The reactor was brought up to full power (3 Mw)
in steps of about 5%, with suitable adjustment of
electrical power to the
increases of reactor power. During steady oper-
ation of the reactor the equipment required only
infrequent and minor adjustments, The system had
a large heat capacity, so temperature fluctuations
were relatively slow. The Leeds & Northrup
PAT controller compensated automatically for
minor changes in heat generation or in heat transfer
conditions. Large increases in heater power were
necessary when the reactor was shut down, but
these were accomplished quickly by using control
settings determined earlier,

loop heaters between

Temperature control was greatly simplified when
it was found that heat removal was adequate
without the use of the air-cocled heat exchanger,
Operation of the heat exchanger was unnecessary
because the fission heat generation was less than
that anticipated. Since fission power generation
could be a controlling parameter in fuel corrosion,
estimates of the fission power were made early in
the operation of the loop by determining the
difference in electrical power required to maintain
the steady operating temperatures with the reactor
at full power and with the reactor off, The differ-
ence in electrical power input was assumed to be
equal to the fission power. Heat from absorption
of reactor gamma rays was assumed to be removed
by the jacket cooling water without affecting the
loop temperatures, and any difference in the
efficiencies of electrical and fission heating was
neglected. The average of several determinations
was 2.8 kw, with a variation between measurements
of +0.03 kw. Later determinations of the fission

27
 

CONCRETE BLOCK SHIELD
6 ft HIGH

 

 

SSD-B-{121A
ORNL-LR-DWG 53424

HB-2

 

A e LITR SHIELD

 

 

400 mr/hr

 

7 mr/hr

 

 

 

 

 

 

HB-3 EQUIPMENT
SHIELD

 

 

 

 

L PUMP

 

 

 

 

%

35 mr/hr

 

 

 

 

 

 

 

 

3-in. PARAFFIN

HIGH-DENSITY BLOCK WALL
' g8ft HIGH

/_,_ CONTROL PANEL

Fig. 23. Horizontal Section Through Loop Shielding and Gamma Field Readings During Operation.

power were made by examinations of the fuel and
of the Inconel loop tubing.

The important operating conditions during in-pile
functioning of the loop are displayed in Fig. 24.
The temperatures plotted are: end of nosepiece
and outlet well of nosepiece (points 24 and 30,
Fig. 9); inlet well of nosepiece (point 29); body
of venturi tube section (point 58), the coolest part
of the system. Except for the well temperatures, it
should be noted that the temperatures are for the
tuke surface, not for the fluid-tube interface,
Thermal insulation and position of heaters had a
large effect on the thermocouples, so that variations

between them were considerable, Although there

28

 

were thermocouples, fluid-tube interface
teinperatures are not known to better than 10 or
15°F.

The discharge side of the nosepiece, as ex-
pected, registered the highest temperatures, which
averaged 1485°F. The well at the inlet of the
nosepiece averaged 1475°F, and the surface of the
venturi ran about 1440°F, Fuel temperature should
have been lowest at the venturi, since all the
out-of-pile heaters and those of the heat exchanger
(It was
necessary to supply some heat in this region, so
this was done at the pump, the component of the
foop with the most heating capability,) The temper-
ature of the tubing adjacent to the venturi was

many

were off or were at reduced power,
TEMPERATURE (°F)

REACTOR POWER (Kw)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   
 

 

 

+Sulnn
ORNL—LR-DWG 16920
1580 — ‘ ‘
| | | |
‘560 _— . — — . — [ - i i —— l ’ -
: i I
| | N
1540 | — | ,

 

 

! ‘ | _~HIGHEST

 

 

 

1520

 

 

 

 

 

 

 

 

1500

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1480

 

1460

 

 

 

 

 

 

1440

 

 

 

 

1420

 

1400

 

 

 

 

1380

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1360
2800
| J l | | | l
2400 'L-—---r e : # : ! P : e .
| ; 1 | _ — 1619
| :
2000 — | — - - — — — - e : - T - . - - - {1545
| i ' i
! | ; } ! j i ; ‘ ! | - 1465
‘ | ‘ ? i ‘ : i | | ! | : 1 | 382 S
1600 [ l e ; e SRR | o ] T C ’-——— ------ ‘ e o | - | 1. s
! | | . | | i | | | ! | | = 1294 §
| | | i ; | ‘ 1 1194
1200 —— - fb —ffi - x : ! - Ce s —_— : R . L f— L) — L ‘ ! — : _ 5
! y : wJ
{> L i ‘ /FUEL FLOW - ; T ! — 1090 2
800 |- M T T = —— Hll. T = : - b ; e R s bt o924
| I | : | ‘ ‘ ‘ ‘ ‘ | ‘ — 0845
a : ’ ! ' ! | !
400 -1 ‘ +— ‘ I — o - T L . - L ‘
\ ‘ ‘ \ | \ | | | ’
o | | Il | J | | l |
0 f 2 3 4 9 10 T 12 13 14 15 16 17 8 19 20 24 22 23 24 25 26 27 28

 

TIME (days)

Fig. 24. Conditions of Operation of Fuel Loop During In-Pile Period.

29

 

 

 

 

 
considered to be a more reliable indication of fuel
temperature than that of the venturi. This tube
temperature, from several thermocouples, averaged
1450°F and was used to estimate the temperature
difference between the hot and the cool zones,
The temperatures are listed with other pertinent
operating data in Table 1. Fuel velocity in the ir-
radiated section of the loop is plotted in the lower
section of Fig. 24. Reactor power, shown in the
center section of Fig. 24, was the standard oper-
ating power of the LITR (3 Mw), with the usual
shutdowns for the various experiments in progress.

The entire loop system, during its operating
period, incidents,
Early in the operation the air activity near the
pump increased sharply. The back-pressure
regulator and the oil trap between the pump and
the charcoal adsorber were found to be obstructed,
The purge-gas pressure built up in the pump was
relieved by opening the bypass valve connecting
the pump and the adsorber, with no interruption
in the operation of the loop,

General air activity from gaseous or volatile
fission products was controlled satisfactorily by
the charcoal adsorter and the reactor off-gas
stack, Gaseous fission products were swept from
the surface of the fuel in the pump by the pump
purge-gas, which was passed through the charcoal
adsorber and then to the reactor off-gas stack,
It was observed that during operation of the loop
the radioactivity of the gas passing through the
off-gas stack was approximately doubled, This
indicated that the loop system was discharging
about 0.001 curie/sec from the charcoal adsorber,
The rate of generation of the fission gases krypton
and xenon, after a day of operation at 3 kw is of
the order of 0.03 curie/sec, If it is assumed that
all these noble gases were removed from the fuel
the charcoal adsorber

functioned with only minor

to the exhaust system,
reduced the activity of the purge gas by an order of
magnitude,

Shutdown of the Loop

After the loop had been operating for 645 hr,
including 475 hr at full reactor power, an emergency
shutdown was produced by a high-temperature
condition at the end of the nosepiece, Circulation
had stopped abruptly, and this circumstance
afforded an incidental observation on the co-
ordination of the loop instrumentation with the
reactor controls for the limitation of temperature

excursions. The temperature rise was almost as

30

rapid as the response rate of the recorder, but the
total rise was not more than 8Q°F before the
reactor was cut off, The record of the potentiometer
which gave the emergency signal and which was
connected to the thermocouple on the hottest point
of the nose piece is shown in Fig. 25, The high-
temperature period is seen to have been of very
short duration.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SSD-A-1173
ORNL-LR-DWG 6545A
1800 7717 71

REACTOR SCRAMMED

AND HEATERS OFF>
;L: FLOW
; STC’)PPED\
% i
!<—[ 1500
o
Ll
o
=
Ll
-

i
HEATERS TUF\’NED CN
1400
0 q 2
TIME {min)
Fig. 25. Temperature Excursion upon Flow Stoppage

in the Loap.

Since the motor drive was outside the permanent
lead shield, it was possible to inspect this
equipment by removal of some of the stacked
shield, The amount of stretching in the rubber
V-belt had been more than the spring-loaded
takeup roller could accommodate, The belt failure
apparently was caused by oil from the air motor
and by the high ambient temperature inside the
shield.

The belt was replaced, and the pump was started
again in an effort to resume circulation of the fuel,

 

 

 

 
 

which had been kept as near as possible to the
1500°F operating temperature, Discharge pressure
was developed by the pump, but no flow took place.
This indicated that some portion of the loop,
probably the heat exchanger, had dropped below the
melting point of the fuel. One of the two heaters
around the heat exchanger had failed early in the
operation of the loop, so this region was poorly
heated. The indicated temperatures of the loop
were slowly raised toward 1600°F in an attempt
to melt out this block. The remaining heater around
the heat exchanger burned out during this effort,
and it was concluded that it would be impossible
to restore flow, All power to the loop was then
cut off, and the fuel was aliowed to freeze.

Removal and Disassembly of the Loop!3

After the loop had been allowed to cool, the
cast lead shields were made accessible by
removal of the stacked shielding and by stripping
off the electrical and instrument leads and service
connections. The pump section was detached as
planned by severing the loop tube bundle between
the reactor shield and the pump shield by means
of the hydraulic shear installed in the utilities
shield. The ends of the fuel tubes were capped
and sealed, and the pump shield, along with the
pump, was moved to a storage location,

With the pump shield removed, the loop with-
drawal shield was moved into position at the
reactor hole, against the matching recess in the
utilities shield. (The withdrawal shield was a
15-ft horizontal cylindrical shield 81/2 in, ID, with

 

laFo_r"_q" more detailed discussion of the disassembly
operation and the disassembly equipment, see Appendix

5 in. of lead cast in an iron shell.) The loop and
jacket assembly was then pulled into the with-
drawal shield by means of cables previously
installed for remote handling of the loop. The
shield containing the loop was moved to the loading
port of the disassembly ‘‘hot cell,’”’ and the loop
was drawn into the cell for dismantling.

The disassembly procedure was dictated chiefly
by the length of the loop assembly and by the
desire to minimize mixing of radioactive fuel with
the insulation, shielding, and other materials of the
loop. The loop was pulled from its jacket (held at
the loading port) as far as the width of the cell
permitted, the extraneous material was stripped
from the fuel tubes, and they were then cut off by
a large horizontal band saw at the loading port,
The withdrawing, stripping, and cutting procedure
was repeated until the loop assembly was reduced
to short lengths of bare tubing. Twenty-two
sections, to serve as metallographic specimens,
were cut from appropriate locations of the loop
tubing., Six additional sections were cut to provide
fuel samples for chemical analysis. The re-
maining tubing was removed to a shielded storage
area to await recovery of the uranium,

The inert atmosphere enclosure around the pump
bow! made this equipment too large for handling in
the band saw. An electric-arc cutting technique
was developed for and the
enclosure was removed from the pump by this
means, The pump bowl was then cut off in the
band saw at the level of the solidified fuel, and
the bowl was stored for uranium recovery, The

remote operation,

remaining part of the pump showed enough radio-
activity to warrant cutting off an additional section
of the bow! for salvage of uranium,

31
 

DISCUSSION OF RESULTS

Measurement of Fission Power Generation

The fission power generated in the loop, as
determined by electrical heating measurements
described earlier, appeared to be about one-third
the anticipated power. Since the fission power
was one of the important parameters of the experi-
ment, every practicable method was carried out to
determine its value. To investigate the pertur-
bation of the reactor flux by the materials of the
loop, a flux measurement in a simulated loop was
An assembly was made which dupli-
cated the arrangement of neutron-cbsorbing ma-
terials in the irradiated section of the loop. The
tubing of this experiment was filled with an alloy
(2 wt % Cd-98 wt % Pb) which had the same
macroscopic absorption cross section as that of
the fuel salt, This assembly, in a water-cooled
jacket, was inserted in hole HB-2. The neutron
activation of the cadmium alloy was measured,

performed,

and the relation between activation and neutron
flux was established by calibrating the alloy with
cobalt foil of known cross section. The value of
the flux found by this method should have produced
over 6 kw instead of the 2.8 kw estimated from
the electrical heat measurements. A possible
cause of the discrepancy was the uncertainty
regarding the position of the loop in the hole,
since the thermal expansion of the loop structure
could not be predicted precisely,

incident at the
surface of the loop tubing was provided by the
activation of cobalt foils which had been attached
at convenient intervals along the loop during
assembly. A better indication of the neutron
flux incident on the fuel was obtained from borings
taken from sections of the Inconel tubing of the
irradiated section, The neutron flux was calcu-
lated from the activity of the borings, using the
data obtained by Bopp'4 for the relation between
The thermal-
neutron flux, as determined by these two methods,
is plotted as a function of distance along the
loop in Fig, 26. The fission power was calcu-
lated, from the flux distribution indicated by the
Inconel activity, to be 2.8 kw, in good agreement
with the electrical heating measurements, The

A measurement of neutrons

the activity of Inconel and flux.

maximum power density at the reactor end of the

 

HC. D. Bopp, Gamma Radiation Induced in Engineer-
ing Materials, ORNL-1371 (April 16, 1953),

32

 

loop, as calculated from this flux distribution,
was 0.4 kw per cubic centimeter of fuel.

Another approach to the fission power generated
in the loop was provided by the activity of the
fuel samples taken from the loop for chemical
analysis. The activity measurements were
carried out by two methods: first, by exami-
nation of unseparated fuel in a gamma-ray
spectrometer, and, second, by measurement of
specific fission products chemically separated
from the bulk of the fuel. In the first procedure,
the height of the Zr-Nb peak in the gamma emission
spectrum was compared with the height of the
peak from a zirconium standard, The Zr-Nb
concentration inthe fuel at the time of measurement
indicated that the fission power had been 2.0 kw.
In the latter determination, zirconium and cesium
were separated from some of the chemical samples,
and the specific activities of each of these fission
products were measured. The activities of two
zirconium samples corresponded to fission rates
of 2.3 and 2.5 kw. In comparing the results of the
zirconium determinations with the 2.8 kw indicated
by electrical heating measurements and by flux
determinations, it should be noted that the al-
lowance for decay of the zirconium fission product
during reactor shutdowns was only approximate,
The value of the fission power from the cesium
samples was 1.5 and 2.1 kw, considerably lower
than determinations by other means, but this value

ORNL—-LR—-DWG 8383A

 

   
  

 

 

NEUTRON FLUX ( neuTrons/cmz sec)

 

— o [ .-
E:‘:___,__ p— T
—_ +

 

 

 

 

 

\

| |
0 2 4 6 8 10 12 14 16 18 20
DISTANCE FROM END OF LOQP (in.)

Fig- 26-
Distance, Determined by Activation,

Thermal-Meutron Flux, as a Functian of
 

was expected to be low because of the escape of
the gaseous parent xenon., From the various
measurements the best estimate of the total fission
power generated in the loop was between 2.4 and
2.8 kw, and the maximum power density was about

0.4 kw/cc,

Determination of the maximum power density
permits a calculation of the dilution factor, the
ratio of the total volume to the irradiated volume,
To avoid confusion over the extent of the neutron
flux, the irradiated volume can be defined as a
volume, generating power at the maximum power
density, which will produce the total power. The
loop was found to have a dilution factor of 180.

Chemical Analysis of Fuel

Tube sections, cut during disassembly of the loop
to provide chemical samples, were taken from both
the in-pile and the out-of-pile parts of the loop to
ensure representative samples, Pieces were also
cut from the portion of the filling line outside the
pump shield for samples of fuel which had not been
irradiated or circulated during operation, The fuel
samples were obtained by collecting borings from
holes drilled in the solidified salt at the ends of
the tube sections. Carbide-tipped bits were used
to prevent contamination from the bits. Saw
cuttings and surface dirt were excluded from the
sample material by drilling a hole with a large bit
before collecting the borings from a smaller bit,
The samples were analyzed for U, Zr, and the
constituents of Inconel (Ni, Cr, and Fe). The
results are presented in Table 3.

The nominal composition of the original fuel
was UF4, 25 mole % (U, 47 wt %); NaF, 62.5
mole %; and ZrF4, 12.5 mole %; uranium enrichment
was about 93%. It is apparent that the preliminary
flushing with Nc:F-ZrF4 (for cleaning and testing)
diluted the UF, in the operating charge of the
loop. During the loop operation, however, the
concentrations of the major constituents do not
appear to have changed. (Burnup of uranium during
operation was only 0.06 g in 4.95 kg of fuel.)
The analytical results show that the chromium
content of the fuel increased as a result of the
corrosion of Inconel and that, probably, the nickel
content decreased during the course of corrosion
reactions, These changes in concentration are
consistent with the observations on unirradiated
loops, 5 although the increase in chromium concen-
tration was two- or threefold greater in the un-
irradiated experiments, These latter loops were
operated at higher wall temperatures and lower
cold-zone temperatures than was the irradiated
loop, and the ratios of the hot-leg surface to the
total volume were different, about 2 in.2/in.3 for
the unirradiated loops and 4 in,2/in.3 for the
irradiated, The higher wall temperatures should
have favored chromium dissolution in the un-
irradiated experiments, but the lower surface-to-

 

15G. M. Adamson and R. S. Crouse, ANP Quar. Prog.
Rep. June 10, 1955, ORNL-1896, p 84; Effect of Heat-
ing Method on Fluoride Corrosion, Examination o
Fluoride Pump Loops 4935-2,4950-1, and 4950-2, ORNL
CF-55-6-99 (June 14, 1955); Examination of Short-
Operating Time Loors Numbers 4695-4 and 4695-5,
ORNL CF-55-7-66 (July 8, 1955); and other CFmemoran-

da in this series,

Toble 3, Chemical Analysis of Fluoride Fuel*

 

 

S l U Zr Ni Cr Fe
ample
P (wt %) (wt %) (ppm) (ppm) (ppm)
Original batch 47.4 83 44 120
of fuel
Filling line 43.2 * 0.5 200 £ 100 10 £ 5 80 £ 10
{control material)
Loop, in-pile 43.7 * 0.2 13.1 £ 0.6 40 £ 10 140 £ 20 120 £ 20
section
Loop, out-of-pile 43.7 1.3 13.4 + 0.4 30 £ 5 150 £ 10 180 * 40
section

 

*The deviotion listed is the average deviation,

33
volume ratio of these loops should have retarded
the accumulation of chromium in the fuel, Ap-
parently, the wall temperatures constituted the
most significant difference in the operating con-
ditions, although the care taken in the preparation
of the salt with the enriched uranium might have
resulted in a mixture containing less corrosive
impurities than were contained in the unenriched
mixtures,

Fission-Product Ruthenium

After operation of the Aircraft Reactor Experi-
ment, deposition of fission-product ruthenium was
noted on the walls of the fuel system,1¢ To
determine the loss of ruthenium from the fuel of
the loop, radiocactivity measurements were made
on two of the chemical samples. The first was
treated to separate the ruthenium from interfering
fission products, after which the activity was
measured. The sample showed less than 5 x 104
d/sec.g, when the ruthenium activity, as calcu-
lated from the fission power, should have been
about 107. The second sample of fuel was not
separated but the gamma activity was measured on
a scintillation spectrometer, both as untreated
solid and as a solution prepared under conditions
not likely to evolve RuO,. The ruthenium peak
was obscured to a certain extent by Compton
scattered radiation from the Zr-Nb peak, so the
fuel was compared with a set of known mixtures
of ruthenium and Zr-Nb. The fuel contained not
over 0.03 times as much ruthenium activity as
Zr-Nb activity, while the ratio calculated (from the
fission rate and decay time) as having been
produced was 0.16. Sections of loop tubing from
which the fuel had been removed were examined
by Robinson and his co-workers.'7 The ratio of
ruthenium activity to that of Zr-Nb on the Inconel
tube walls was found to be about 5, whereas the
ratio should have been only 0.6 if the deposits
on the walls had been traces of fuel. Evidently,
most of the ruthenium produced in the fuel plated
out on the walls of the containing tubing during
operation, as was observed in the ARE.

Metallographic Examination

To facilitate handling and polishing, the fuel
was removed from the tube sections that were cut

 

. 1. Robinson, S. A, Reynolds, and H. W, Wright,
ANP Quar. Prog. Rep. March 10, 1955, ORNL-1864, p 13.

7M. T. Robinson and T. H. Handley, ANP Quar.
Prog. Rep. June 10, 1955, ORNL-18%6, p 167.

34

 

from the loop for metallographic specimens
(described in Appendix A). Specimens were cut to
convenient size, were mounted and polished, and
were etched electrolytically in o 10% sulfuric
acid bath, following standard
procedures for Inconel. The metallographic
preparation and examination were performed by the
Remote Metallography Group of the Solid State
Division.'® The locations from which specimens
were taken are shown by heavy lines on a sketch
of the loop in Fig. 27. Control specimens ~ pieces
of as-received tubing cut from the ends of the
tubes used to fabricate the loop — were examined
for comparison with the samples from the loop.
Some of the control specimens for the nosepiece
were also subjected to the same bending and
treatment which this
received in fabrication.

The heat treatment of the material from which
the metallographic specimens were taken was just
that received in operating the loop, with the
exception of the nosepiece (Figs. 37 through 40)
and the heat exchanger. These components were
put through brazing operations during fabrication
of the loop. The brazing treatment involved
holding of the parts at 1950 to 2050°F for ]/2 to

1 hr, followed by forced-air cooling.

remote-handling

brazing loop component

In general, the corrosive attack observed was
slight, and any metallurgical changes were insig-
nificant, Corrosive penetration of the specimens
from the loop is indicated in Fig. 27; also given
are the locations of the specimens shown in
Figs. 28 through 40. First, the unirradiated portion
of the loop will be considered; Figs. 28, 29, and 30
present typical specimens from the 2-ft section
of the loop between the reactor shield and the
pump, This region, along with the continuous heat
exchanger, was the only portion of the loop un-
heated during actual irradiation, The region was
maintained at about the temperature of the rest of
the loop by the molten salt flowing through it.
The specimen shown in Fig. 28 was cut from the
inlet leg to the reactor; the specimens shown in
Figs. 29 and 30 were taken from the outlet leg,
This part of the outlet leg (Figs. 29 and 30)
showed the minimum corrosive attack, an average
penetration of less than 0.5 mil and maximum
penetration of only 0.5 mil, while the inlet leg,
Fig. 28, also showed comparatively light cor-

rosion. No deposits of mass-transferred material

 

8. J. Feldman ef al,, Metallographic Analysis of
Fuel Loop 11, ORNL CF-55-6-22 (June 21, 1955).
 

*U0I404AUB ] FAISOMOD) PUD ‘SuUldwydad§ Jo suolipao] ‘Buiqn} jo syibua- uaamiag siuior Buimoys doo jang jo ys4ay§ */LT *B61 4

{4}) 4007 40 ON3 WOH4 JONVLSIO

 

 

 

 

 

 

 

 

 

 

 

 

 

 

trh < 2 i Ol 6 8 L 9 g b ¢ z b 0
[ I _ I _ I _ I f [ i | 4 _ _
SpW NI NGILYY L INTD 40 HLdI0 LNISIHLIY SHINAN
x>m\§moo? be 014
XYW G0 XYW Qb . te b4 XvYW 80
AYS O AvGo 8¢ 9 : AY S0
e g1y XTW O XUW G} oo og XTWOT
AY GO AT B0 AY G0 /
\ 1T
v / \ N
- \r—\k.\%\ r\J / . —
T = I T —= 1 gasti |
v 914
xomeo 7 - OID-5F 04
AV SO S 2b Ol
il [ —_— N I [ - ] \\H 1]
XYW B O)
XYW S0 : XY S0 : XN G0 XYW S0 XYW B0 . XYW G 2 VA8 8% 914
AvGo< ©¢ 9 AT G O< OF VM AW GO AT GO AV S0 S8 93 Ry5T AV S0
9¢ b1 ot o114 x\,,qq...%ol,

rPBER OMO—UT-TINHO

35

 
UNCLASSIFIED
RMG 1100

 

Fige 28, Inconel, Unirradioted, Exposed to Fluoride Fuel ot About 1450°F for 645 hr. Etchant: electrolytic

HZSU“. 250X. (Secret with caption)

UNCLASSIFIED
RMG 1154

 

Fig. 29. Inconel, Unirradiated, Exposed to Fluoride Fuel at About 1460°F for 645 hr. Etchant: electrolytic

HISDd. 250X. (Secret with captieon)

36

 
 

H

 

H

2

2

 

Fig. 30,
504. 250X.

Fig. 31.

SO,. 250X.

UNCLASSIFIED
RMG 1152

Inconel, Unirradiated, Exposed to Fluoride Fuel ot About 1460°F for 645 hr. Etchant: electrolytic

(Secret with caption)

‘-‘ UNCLASSIFIED
L o RMG 1104

Inconel, Unirradioted, Exposed to Fluoride Fuel at About 1450°F for 645 hr. Etchont: electrolytic

(Secret with caption)

37

 
were found in this or in any other part of the loop,
The apparent layer in Figs, 29, 30, and certain
other specimens is an oil stain produced during
etching
void between the specimen and the plastic mount,

and was caused by material from the

Figure 31 is a photomicrograph of a longitudinal
This

section, which was exposed to various velocities,

section of the entrance to the venturi throat,

shows no unusual corrosion or erosion,

A sample, from the inlet leg within the reactor
shield, which is typical of the entire unirradiated
part of the loop is pictured in Fig. 32. Maximum
penetration was 1 mil; the average was 0.5 mil,
Another also unirradiated, from the
same piece of tubing is shown in Fig. 33. This
sample exhibited penetration at a crevice to a
depth of 4 mils, although the average penetration
was 0.5 mil, and the general corrosive attack was
about the saome as that observed in the earlier
specimens. [he os-received control specimen for
this length of tubing, Fig. 34, showed a crack,
similar to the crevice in Fig. 33, which must have

specimen,

manufacture of the tubing.

been introduced in

 

Fig. 32.

H, 50

250, 250X. (Secret with caption)

38

in other
specimens (in both as-received controls and loop

Such cracks were found occasionally

specimens from this lot of tubing) without any
apparent difference in frequency. [t was concluded
that the occasional voids to a depth of 3 or 4 mils
were not results of corrosive penetration to this
depth but were formed by penetration of 0.5 to |
mil at an existing crack present in the original
tubing. This view is substantiated by the absence
of such crevices in the
irradiated section, which was constructed of tubing

different from that used in the unirradiated parts,

specimens from the

The area of maximum corrosion (Fig. 35) was
found in the outlet leg between the irradiated zone
and the heat exchanger: average penetration 1 mil
and maximum penetration 2.5 mils. Mo reason
could be found for the increased corrosion at this
point, The control specimen from this end of the
length of tubing, Fig. 36, did not appear different
from the others. The chemical analysis (see
Appendix B) was not consequentially different from
that of the other tubes.

that fission

There is the possibility
heating in the irradiated section

UNCLASSIFIED
RMG 1112

Inconel, Unirradioted, Exposed to Fluoride Fuel at About 1460°F for 645 hr. Etchant: electrolytic

 
 

UNCLASSIFIED
RMG 1116

 

Fig. 33, Inconel, Unirradiated, Exposed to Fluoride Fuel at About 1470°F for 645 hr. Etchant: electralytic

|2504. 250X. (Secret with caption)

 

Fig. 34, Inconel, As-received Control Specimen. Etchant: electrolytic HESD". 250X, (Secret with caption)

39

 

 
 

 

UNCLASSIFIED
RMG 1205

-

Fig. 35. Inconel, Unirradiated, Exposed to Fluoride Fuel at About 1485°C for 645 hr. Etchant: electrolytic

H.50,. 250X. (Secret with caption)

2

 

40

4

UNCLASSIFIED
RMG 1138

Fig. 36. Inconel, As-received Control Specimen. Etchant: electrolytic H2504' 250X. (Secret with caption)

 

 
 

UNCLASSIFIED
RMG 1126

 

Fige 37. Inconel, Heated to About 2000°F for 1&—-1 hr, Irradioted and Exposed to Fluoride Fuel ot About
1485°F for 645 hr. Etchant: electrolytic HQSDA' 250X. (Secret with caption)

  
   

UNCLASSIFIED
RMG 1130 -

o1 :

® - w

Fig. 38, Inconel, Heaoted to About 2000°F for 15—-1 hr, Irradioted and Exposed to Fluoride Fuel at About
1485°F for 645 hr. Etchant: electrolytic H2504. 250X. (Secret with caption)

41

 
UNCLASSIFIED
RMG 1132

 

- ¥§ s 5
- e . i . o
. * i fo e A -
. 3 e - i .
% . . -~ ¢ o g, P
B 5 - - - TS _‘_‘f . 4F -

Fig. 39. |Inconel, Heated to About 2000°F for ],3-] hr, lrradiated and Exposed to Fluoride Fuel at About

1485°F for 645 hr. Etchant: electrolytic HZSDA. 250X. (Secret with caption)

 

| Yt UNCLASSIF IED
B AT % RMG 1134
oy -\'m s -'r.f‘v-h\"‘ v
# ; “'l:. ;‘- ¥ o F
.? s + - X & N
l‘N " ._.\ ' :
- . ¥ = -I“\'\J‘_ - _. D
-y & . N o
- . . .-.‘ d s g -
. I. & e ?.. o I".
>z ’-i.4 rL e : o A
; LN
.. . ‘m r“. '_ “od 5, ni
i -\"‘l . ' 5".“ * .‘!..‘-‘ ‘
A e ‘.} . T \“i“" -‘ i
R = 2 § h= - £ e - :‘-u.‘ e
¢ b e gl Sy Ay
o AL e -8
k .'.. & o . - e
2 4 qu - - =
. ? gt < i R e
. 4 . ,". ‘ o * !'
» " » - r'. .
- r -
- * + &
5 c - ‘: o i
‘ 2 i
: ey f
' " S Jee T3
o Ji- @ % 9y 'II
v\ "%
? A~ —
- -:f . T ""_Fr— '\
- r‘ll 3 '. '.- \
» Gt i - " .. \
‘_I. .o g \ |

Fig. 40. |Inconel, Heated to About 2000°F for 1-2’_1 hr, lrradiated and Exposed to Fluoride Fuel at About
1485°F for 645 hr. Etchant: electrolytic HESDd. 250X%. (Secret with caption)

42

 
 

raised the temperature of the fuel in the outlet leg,
thereby accelerating corrosion in this part of the
loop. On the other hond, the outlet portion of
the irradiated section does not show oppreciably
greater corrosion than does the central region.
Therefore, a better
temperature differences due to variations in contact
with the heater or that
chemical composition of the tubing accounts for
the accelerated corrosion.

The specimens from the irradiated
(nosepiece) showed light
attack, giving no evidence for acceleration of

explanation is that local

nonuniformity in the

section
relatively corrosive
corrosion by irradiation. The specimen in the
most intense [lux is shown in Fig. 37; average
penetration was 0.5 mil, and maximum penetration
The same extent of attack
was found in longitudinal sections of a specimen
in a flux almost as high; see Figs, 38 and 39.
The former shows the outer wall in the U-bend,
and the other the inner wall. No difference in
corrosion was observed in the two walls, Slightly

was about the same.

 

Fig. 41,
(Secret with caption)

Inconel, As—eceived, Thickwalled Control Specimen,

more severe corrosion was found in o specimen
from the downstream side of the irradiated section,
Fig. 40, which was exposed to a flux about half
Penetration, still only light,
was to an average depth of 0.5 mil, with an
occasional void to 1 mil. This part of the ir-
radiated section also showed a larger grain size
Appreciable
grain growth occurred in the irradiated section
during the brazing operation in fabrication of the
assembly.
brazing (Figs. 41 and 42, respectively), show the
change in grain size. The uniformity of conditions
within the brazing furnace was not known. As a
whole, the irradiated section showed no corrosion
or metallurgical effects different from the rest of

the maximum level.

for some caouse not determined,

Control specimens, before and after

the loop.

In summary, the metallographic examination
showed only light general corrosion. Void for-
mation had penetrated to depths of 2 to 3 mils,
but the area of maximum penetration or heaviest

corrosion was not found in the irradiated section.

UNCLASSIFIED
RMG 1218

Etchant: electrolytic oxalic acid. 250X.

43

 

 
 

 

UNCLASSIFIED
RMG 1219

Fig. 42, Inconei, Thick-walled Control Specimen Heated to About 2000°F for 145+1 hr, Etchant: electrolytic

oxalic acid, 250X. (Secret with caption)

Voids associated with surface cracks were found
in the unirradiated part of the loop to depths of
4 mils, but similar cracks were observed with the
same frequency in the as-received control samples,
It was concluded that the voids associated with
cracks did not represent corrosive penetration to
4 mils but that maximum penetration was only
2 to 3 mils. No evidence of mass transfer was
found, nor was any enhancement of corrosion by
irradiation observed. This latter observation is
in agreement with the conclusions indicated by
some static capsule irradiations, 1920

The corrosive attack in the irradiated loop may
be compared with that in the unirradiated loops?!
which were operated at about the same upper fuel
temperature (1500°F) but at a much higher wall

 

19 . s. Richardsen, D. C. Vreeland, and W, D. Manly,
Corresion by Molten Fluorides, ORMNL-1491 (March
17, 1953).

0y, v, Goeddel, Cyclutron lrradiation of Fused
Fluorides in Inconel at Elcvated Temperatures, MAA-

SR-208 (Jan. 26, 1953).
2150e ref 15, p 33.

44

 

temperature (1560 to 1700°F) and a much lower
cold-zone temperature (1200 and 1300°F, as
compared with 1440 or 1460°F for the irradiated
loop), These differences in operating conditions
would tend to accelerate mass-transfer processes
and dissolution of chromium from the hot zone
in the unirradiated loops. Another difference in
the loops which would favor void formation and
removal of chromium in the hot zone of the un-
irradioted loops was the lower ratio of hot-leg
surface to total volume, about 2 in.2/in.? for the
unirradiated, and 4 for the irradiated loop. In
accordance with these differences in conditions,
the corrosive attack in the irradiated loop was
Corrosive
penetration was 2 to 3 mils for the irradiated
loop and 5 mils for the unirradiated loops with
wall temperatures under 1600°F but was up to
25 mils in loops with higher wall temperatures,
Mass
loops with the more severe conditions but was

much lighter than in the unirradiated.

transfer was observed in the unirradiated

not observed in those with wall temperatures

under 1600°F; also, it was not observed in the
R A e

irradiated loop. |t can be concluded that the
differences observed between the unirradiated
and the irradiated loops disclosed by metallo-
graphic examination are those to be expected from
the less severe temperature conditions of the
irradiated loop.

Conclusions

The fission power generation during operation of
the loop was between 2.5 and 2.8 kw, and the
maximum power density was 0.4 kw/cc. Chemical
analysis of the fuel showed no decomposition by
irradiation, Fission-preduct ruthenium was largely
deposited from the fuel onto the tube walls, The
increase in chromium and the slight decrease in

nickel impurities in the fuel are consistent with
the findings in unirradiated {oops, although the
increase in chromium was not nearly so great,
probably because of the higher wall temperatures
loops.  Metallographic ex-
amination of the loop revealed a general, rela-
tively light corrosive attack. Average penetration —
in the usual form of subsurface voids — was 0.5
mil; maximum penetration was 2 to 3 mils. Cor-
rosion was lighter than that observed in unirradi-
ated loops, undoubtedly because of the practically
isothermal operation of the irradiated loop and
because of its more moderate hot-zone wall temper-
atures. There was no acceleration of corrosion by
irradiation,

in the unirradiated

 

Appendix A22

DISASSEMBLY PROCEDURE AND EQUIPMENT

The size of the radioactive parts, the variety
of materials, and the accountability problem made
the disassembly operation a unique one in remote
For this reason, as well as to provide
an understanding of the form and the condition of
the samples examined, a thorough discussion of
in the

handling.

the equipment and operations involved
disassembly is included in this report,
The constraints placed on the irradiated appa-
ratus by the operating conditions of the experiment
precluded any extensive provisions for disassembly
in the apparatus itself, The size of the loop
assembly rendered unsuitable most of the existing
remote-handling equipment, so that special devices
had to be obtained.
a policy of modifying existing equipment or of

In procuring this equipment

revising readily available stock models was
followed,

The problem presented by the length of the
loop, compared with the dimensions of the remote-
handling (hot) cell, was solved by cutting a loading
port in the rear wall of the cell. The port was
closed by means of a hydraulic door and was
recessed so that there would be no stray radiation
beams when the withdrawal shield for the loop
was in position, Wire cables through both the

shield and the cell enabled the loop and its jacket

 

227he authors of Appendix A are R. M. Carroll and
W. W. Parkinson.

to be moved in either the shield or the cell, Three
C-clamps having hydraulic plungers were made in
various sizes {maximum openings, 1 to 8 in.)
All the
equipment was operated by, or in conjunction with,
the electric manipulators which are standard
equipment for the larger hot cells and which have
been described elsewhere.?3

Cutting devices played the most important part
in the disassembly operation and required the
most developmental effort, A horizontal band saw
(Fig. A.1) of 8- by 10-in. capacity was set up just
inside the loading port for major cutting operations.
This band saw was fitted with a hydraulic vise

to hardle any part of the loop or jacket.

and was modified for remote operation. Vacuum
cleaners having filters of the replaceable paper-
bag type were used to collect cuttings from the
saw and to prevent the spreading of fuel or other
radioactive materials, The suction of the vacuum
cleaners substituted to some degree for the
conventional liquid-cooling system which was
removed from the saw. The band saw proved
very versatile and dependable, An electric hack
saw (Fig. A.2) was adapted for cutting small
pieces inaccessible to the band saw, It was
carried by the manipulator in place of the standard

manipulator tongs and was fixed to its own vise

 

235, E. Dismuke, Physics of Solids Institute Quar.
Prog. Rep. July 31, 1951, ORNL-1128, p 6.

45

 
UNCLASSIFIED
. PHOTO 137&6

. B i1
; MANIPULATOR '

   
 
 
  

   

WINDOW

 

BAND SAW

C,&RFE'!EFI‘ SHIELD

‘h

h , \

VACUUM  \ ; g A winoow
CLEANER -

;"-I- ‘\t L P 3
E ——— - —

| wasT ) | '
1 & STORAGE SHIELD
Al -

   

 

CAN

Fig. A.1. Equipment for Disassembling the Loop in the Hot Cell.

46
      
   
  

; B o ¥ 1.
" UNCLASSIFIED
. y  PHOTO 15092
| G &
' = 5
” e
— I

D,'...'w' 'mflm- VoL,
w N }
o .‘Tm\u il - r-

e ——
—

  

Fig. A.2. Remotely Operated Hack Saw.

A pair
shears was

by means of a hydraulic feed mechanism,
of ‘“‘hawk-bill’" sheet-metal hand
adapted for remote hydraulic operation so that
they could be handled by means of special manipu-
These shears (Fig. A.3) proved
useful for cutting a variety of materials, including
thermal

lator tongs.
insulation and hose clamps holding the
heaters to the loop.

An additional item of cutting equipment was
required for dismantling the pump. The pump with
the stainless steel sheet enclosure around the
bowl was too big for the band saw, and the sheet
metal was too thick for the shears. An electric-arc
cutting apparatus consisting of a 300-amp d-c arc
high-frequency-arc starting at-
tachment and cutting electrodes was found to be
satisfactory when adapted for remote operation by
the hot-cell manipulator,

welder with a

An electric impact wrench was required specifi-
cally for removal of the flange that closed the
water jacket around the loop. The impact wrench
carried by the manipulator was used with various
sockets and extension rods to remove the cap
screws holding the flange. Furnaces, containers,
and miscellaneous items, the arrangement of which
is shown in Fig. A.1, made up the balance of the
hot-cell equipment,

The first step in the dismantling operation was
the removal of the water-jacket closure flange
with the impact wrench (Fig. A.4). The loop was
then withdrawn a few inches from the jacket,
which was held at the loading port by the band-saw

vise, The loop support rods were cut with the
band saw without disturbing the fuel tubes.
Instrument and electrical leads, which passed

through Kovar seals in the flange, were cut with

47
 

48

Al

  

fi- PHOTO 139437
— . .

 

UMCLASEIFIED

        
 

the hydraulic shears in order to permit removal of
the flange from the loop tubing.

The disassembly of the loop proceeded from
this stage in steps: first, the loop was withdrawn
from the water jacket as far as the width of the hot
cell permitted; second, insulation, shielding,
leads, and heaters were stripped from the exposed
section; finally, the bare fuel tubes and support
rods were sawed off to permit another withdrawal.
Sprcimen sections were cut from the fuel tubes
for metallographic examination and to provide fuel
for chemical analysis. The remaining fuel tubing
was stored for future recovery of the uranium,

The locations from which the metallographic
specimens were cut are indicated in Fig. 27.
These specimen sections were marked with saw
cuts for identification before melting out of the
fuel, They were held vertically in a stainless
steel cylinder having a wire-mesh bottom, This
holder rested on lugs in a stainless steel furnace
pot, with space beneath the holder for collecting
molten fuel, A close-fitting lid on the furnace pot
was equipped with a gas connection and a thermo-
couple well. To melt the fuel out of the metallo-
graphic specimens, the furnace was held at 800°C
for about half an hour and was continually purged

UNCLASSIFIED
PHOTO 13884

B ’ R

HE

 

Fig. A.4. Removal of Closure Flange.

 
with argon. The specimens were then ready for
sectioning, mounting, and polishing.

The tube sections for chemical samples (~2 in,
long) were gripped in the chuck of a drill press
set up in one of the hot cells having a master-
slave manipulator, Carbide-tipped bits, to elimi-
nate contamination of the sample with bit ma-
terial, were mounted inside plastic cups located
below the fuel tube, Fuel was bored from the end
of the tube and collected in the cup. The samples
required were small enough so that weighing
could be done by hand, but dissolving and
dilution were done remotely,

The remaining part of the loop system, the pump,
was lowered into the hot cell in its own operating
shield. One wall of the pump shield was movable,
and access to the pump was gained by hoisting
away this wall. The stainless steel enclosure
around the pump bowl was cut away with the
arc cutting equipment described earlier. After
removal of the enclosure the pump bowl was
stripped of heaters and thermal insulation, and the
inlet and discharge fuel tubes were sawed off with
the electric hack saw (Fig. A.5). The pump could
then be mounted in the band-saw vise, and the
bowl was cut off just cbove the level of the
solidified fuel. Enough radioactivity remained
in the upper part of the pump to warrant cutting a
second portion of the bowl from the superstructure
of the pump. The pump bowl parts were stored
with the loop fuel tubing, pending salvage of the
uranium,

The usefulness of large storage shields inside
the hot cell was emphasized by the necessity for
replacing a bhroken band-saw blade. The pump
and radicactive parts of the loop were placed in
shields provided in the hot cell, and the radiation
level was found to be low enough to permit entry
into the cell for replacement of the blade. The
failure of one of the manipulators was handled in

WRCLALUFIED
Hd 0 a1

 

Fig:. A.5. Pump After Remote Stripping.

the same manner and emphasized the desirability
of having two manipulators available,

In conclusion it can be said that the equipment
and methods described above worked acceptably,
A total of 440 hr was required for the disassembly,
during most of which two men were available,
The greatest reduction in the time consumed would
have been accomplished by use of a thermal
insulation stripper, probably a small rotary saw
operable in the manipulator in restricted areas,

49

 
Appendix B

ANALYSIS OF INCONEL TUBING

(As-received tubing before fabrication)

 

Analysis (wt %) of Specimens

 

 

Element For Irradiated For Inlet Tube For Outlet Tube Method
Section to Reactor from Reactor

Ni 76.2 74.9 74.8 Gravimetric
Cr 13.5 15,2 15.2 Volumetric
Fe 6.4 7.1 7.0 Volumetric
Mn 0.23 0.13 0.21 Colorimetric
Cu 0.09 0.28 0.26 Polarographic
Si 0.25 0.15 0.12 Gravimetric
Al 0.064 0.064 0.057 Spectrographic
Ti 0.224 0.169 0.174 Spectrographic
S 0.005 0.003 0.006 Wet method
C 0.651 0.043 0.027 Wet method