r
Di
*
.
.
.
.
.
od
-
-
*
I OFI
ORNL
P3188
:
.
.
.
_
L
.
.
.
.
.
.
"
.
.
.
.
:
.
.
NA
| 1.25 1.4 11.6
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
rit
ORNL f 3188
CONF-670916-01
+P
MASTER
GERO 2013
AUG 2 2 1969
.C.$3.06. ww.65
TECHNIQUES FOR STABILIZING LIQUID METAL POOL BOILING
R. E. MacPherson
Oak Ridge National Laboratory
Oak Ridge, Tennessee
rm
S
Y
marice
cance
ABSTRACT- Liquid-Detal-cooled reactor projects at the Oak Ridge National
Laboratory have required the operation of potassium pool boilers for materials
compatibility tests and for performance a:id endurance tests of coinponents. In
many diverse facilities, problems with superheating of liquid metal have been
encountered. To avoid these problems, special operating techniques and mechani-
cal devices have been developed. Vapor traps in the form of metal sleeves me-
chanically attached to heat transinitting surfaces in the boiler have been
enco
S
me
effective in stabilizing boiling. Independently heated devices designed to dis-
charge vapor il'o the boiler pool serve to initiate the boiling process smoothly.
Related experience has come from tests of the cavitation characteristics of
liquid metal pumps, wherein the presence or absence of inert gas has affected
cavitation inception, a form of boiling.
Research sponsored by the U.S. Atomic Energy Commission unier con-
tract with the Union Carbide Corporation.
DISTRIBUTION OF THIS DOCUMENT IS INMITEN
INTRODUCTION
Liquiu-metal-cooled reactor projects at the Oak Ridge National Laboratory
have required the operation of potassium pool boilers for materials compatibility
tests, for heat transfer investigations, and for performance and endurance tests
of components. Many diverse test facilities have been operated ranging from re-
fluxing capsules through natural circulation loops to very sophisticated forced
cilni?alid on systems. A common characteristic in most of the early boilers was
the occurrence of liquid metal superheating and explosive boiling, leading to
thermal shocks, operating instabilities, and, in some cases, equipment failures.
The phenomenon was not understood initially, and considerable effort was neces-
OS
fo
So
sary to diagnose and to characterize the superheating process.
It became
evideint
that techniques were needed to provide stable nucleation sites in potassium pool
es were
boilers. Nucleation techniques and devices were developed and were found to be
sufficiently effective that no difficulties with explosive boiling were en-
countered in the systems built after 1963.
Avoiding pool superheating and developing devices for stabilizing boiling
involves two related requirements. The first requirement is to initiate boiling
in the pool, and the second requireinent is to maintain small stable vapor pockets
to provide nucleation sites after boiling has begun. Theoretical and experimental
work has been carried out to describe the relationship between bubble radius, or
hole radius, and the superheat required to maintain a stable liquid-vapor inter-
face. ! 2 3 This work has shown that stable boiling, once initiated, can be main-
ei
tained with only a modest ainount of liquid superheating (i.6., 10-25°F) by the
inclusion in the boiling system of appropriate nucleation sites such as drilled
holes. However, it has also been demonstrated that in a clean, degassed system,
e've
LEGAL NOTICE
This report was prepared as an account of Government spor.sored work. Neither the United
States, nor the Commission, nor any person acting on behalf of the Commission:
A. Makec any warranty or repis ntation, expressed or implied, with respect to the accu-
racy, completer e88, or usehulness of the information contained in this report, or that the use
of any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rigats; or
B. Assumes any ilašilities with respect to the use of, or for damages resulting from the
use of any information, apparatus, method, or process disciosed in this report.
As used in the above, “per son acting on behr of the Commission" Includes any one-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
Buch employee or contractor or the Commission, or emplo;ee of such contractor prepares,
disseminates, or provides access to, any Irformation pursuant to his employment or coriact
with the Commission, or his employment with such contractor.
Fig. !
UNCLASSIFIED
ORNL-LR-DWG 75772

-END.CAP
-METALLIC PLUG
Mgo
-HEATER SHEATH
3/16 TYPICAL
- HEATING ELEMENT (4)
--- HEATER LEADS
NOTE: DIMENSIONS ARE
IN INCHES
NOT TO SCALE
LAVA PLUG
Watlow Firerod.
Rinn
: no:
UV02
VOLVOVOU
.
.

UNCLASSIFIED
ORNL·LR--DWG 75773

-VACUUM LINE
WHET
-AIR JACKET
TALEZ
AIR INLET
EMUA
ZN
f
- THERMOCOUPLE
a
S
WELL
.
.
-BAFFLE
J-FILL LINE
r
MZUINITIALITZAT
Ausricissussusa
AULUSSASSI
LUOTETTIIURE
u
-
HEATER

0
3
susisus
Cyruusta
Luuumiga
1 2
27
INCHES
Kich
DRAIN LINE
Heater Test Capsule.
Fig. 2
..
5
:

осоо o o o o o o o o
OOOOOOOOOOO
ORNL-DWG 65-748
-TUBE
0.017 OD
0.011 ID
BUILT-IN
"HOT FINGER"
onone; OOOOO
0:0 0 0 0 0 0 0 0 0 0 0
SITTIIIIIIIIIITRIT SIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
..
.
0.0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.....
CHAMFERED
SLEEVE
Nucleation Sites
ооооооооооооооооооооооооооо
III/
T
.
T
KA
IZ
VE
.
.
SHRINK-WELDED
SLEEVE

sodium can be superheated over 600°F without boiling initiation, 3 Most of our
experience with spontaneous initiation of boiling in typical liquid metal
facilities falls between these extremes. Although stable liquid-vapor inter-
faces have been sustained at low values of superheat, it has not been typical
for boiling to start in a potassium pool contained in a boiler having "as-
received" surfaces until superheat values of 100-200°F are attained. With
pool boilers, superheating of this magnitude can result in large amounts of
mo
stored energy available for rapid release with resultant mechanical and thermal
shock effects.
Work has been done on nucleation devices to satisfy both the requirement
to initiate boiling in a controlled fashion and the requirement to sustain
stable boiling. These devices are best described in the framework of the test
1
programs for which they were developed ,
ELECTRICAI. CARTRIDGE HEATER TESTS
During a development program to provide cartridge heaters (Fig. 1) capable
of sustained operation at 1540°F with heat flux values of 120,000 to 150,000
Btu/hr ft? (40-50 watts/cm2), 85,000 hr of high temperature testing were ac-
cumulated in refluxing potassium capsules (Fig. 2). In early tests, heater
failures occurred as a result of high temperature transients during the initial
of the heater; and at heat flux levels abune 60,000 biu/hr ft2 (~20 vatts/cm2),
a sufficiently large temperature gradient could be generated between the potas-
sium pool and the heater surface during the transient to preclude any subse-
quent return to a nucleate boiling condition. Rapid heater temperature rise
and failure of the heater would result. At heat flux levels below 60,000
Btu/hr ft, it was generally possible for the liquid potassium to reestablish
contact with the heater surface; and another superheat-flash cycle would occur.
This latter circumstance provided a mechanism, although admittedly somewhat
necessary to provide some means to sustain vapor pockets in the system after
they had been formed during the inician iiasiing transient. Figure 3 illus-
Lote contre las
trates three devices chat were used for this purpose.
The slain sleeve and the chamfered sleeve served as heat dams mechanically
ser
attached to the heater surface by shrink-velding. The heat flux through the
plain shrink-welded sleeve generated a temperature drop of approximately 50°7
(for .032-41. wall at 150,000 Btu/hr. ft?), which was sufficient to maintain a
stable vapor pocket once vayor. had been produced in the pool by some other
means (normally flashing).*
The chanfered shrink-welded sleeve was designed
to trap vapor cuore efficientiy and to produce a higher temperature difference
between the pool and the heater owing to the presence of a discrete, low-
conductivity vapor pocket.
The third nucleation device shown in Fig. 3 is a thin-talled tube installed
so that it penetrated the heater along a diameter between the two lower heater
coils. Because the temperature gradient between the heater coil and the outer
was
SIC
surface was quite steep owing to the presence of electrical insulation, the
center portion of the to be tended to run from 50-150°F hotter than the heater
It is apparent that, dependent on the nature of the fit between the sleeve
and the lieater surface, there is some unspecified additional temperature gradi.
ent hetween the inner sleeve surface and the heater surface when vapor exists
in the annulus.
and the hea terasurface, there some unspecified additional te
surface for the normal operating power ranges.
This temperature gradient was
more than sufficient to stabilize within the tube a liquid-varor interface that
served as a nucleation site,
m
The relative performance of these three parasitic nucleation sites is shown
by Table 1. The performance criterion was the minimum temperature to which the
an
was
pool could be cooled while still sustaining stable boiling. Since the test
tuchnique
sed lowering the pool temperature t a constant power inpuć until
a super heating-flashing event occurred, it was necessary to restrict testing to
the low power range (<20 watts/cm2) where teater failure was not a problem, as
previously discussed. The general conclusion was that the nucleation performance
appeared to be a direct function of the imposed temperature difference. The
simple shrink-welded sleeve vás the poorest performer, with the performance im-
HANCE
pruving as heat flux was increased. The chamfered sleeve of the same thickness
gave somewhat better results owing to the insulating effect of the vapor pocket,
res
and its performance improved as both the thickness and the heat flux leve). were in-
creased.
The tube showed the same general level and trend in performance as the
thicker chamfered sleeve,
Table 1. Minimun Temperature for Sustained Stable Boiling
Heat Flux
Nucleation Site
20' «atts/cm2
(°F)
10 watts/cin?
(°F)
Sleeve (1/32 in, thick)
1350
1575
Chamfered sleeve
1/32 in. thick
1125
1300
1/16 in. thick
950
1075
Through-tube
900
1125
The primary application of these devices required long-tern use at a pool
temperature between 1500-1600°F, and the potassiun-structural metal (type 304
stainless steel) compatibility problem at the hot spots vas demonstrated to
be too scvere in the application of both the chamfered sleeve and the tube.
Short-term failures were experienced in both cases. For this reason, the
simple sleeve was chosen as the standard vapor trapping nucleation site.
the well. ver.wn technique of incozporating a porcelain chip in a laboratory
distillation rig to promote sniooth boiling prompted tests of high! purity
beryllia and alumina as nucleation sites. Initial performance was, in every
case, completely satisfactory. Boiling was initiated snoolhly on startup and
maintained staily during system hcatup. As might be expected, less satisfactory
performance was attained during a second reheat cycle. Apparently thermal ex-
pansion effects and gas solubility effects resulted in vetting and flooding of
the pores of the ceramic test pieces, thus destroying their effectiveness as
nucleation sites.
BOILING POTASSIUM MATERIALS COMPATIBILITY TESTS
An extensive series of tests has been conducted to study the erosion and
corrosion of high temperature structural materials by high velocity potassium
vapor-liquid mixtures. Two-phase, forced-circulation loops have been operated
in this program at boiler outlet temperatures from 1600 to 2200°F for a moral
of 26,000 hr. Figure 4 illustrates a typical test facility,
Stable boiling is a primary criterion for control of these long-term
metallurgical test facilities. One austenitic stainless steel system recently
completed a 10,000-hr run at a boiler outlet temperature of 1600°F. During
ORAL-DWG 63-3328
TEST SECTIONS
DRYER
SURGE
TAIK
CONDENSER-
-BOILER
15-% ZE ALLOY TO
TYPE 3:6 STAINLESS
STEEL JUX'CTION
U
MWuenar:
·
Nok
COOLER
ZIRCONIUM
UHOT TRAP
--PREHEATER
AIR-V
10

u
King
UL
MORY
-CONTROL VALVE
FARADAY PUMP
- FLOWMETER
Nb-1% ZT ALLOY TO
TYPE 316 STAINLESS
STEEL JUNCTION -
WATER-COOLED
O! TRAP
ELECTROMAGNETIC
PUMP
12-in. VACUUM BLOCK VALVE -
LIQUID-ilige
COOLED
ТКАР
FLOWMETER
- MECHANICAL VACUUM PUMP
-VACUUM CHAIVISER
10-in.
DIFFUSION
PUMP
4-in. DIFFUSION
PUN:P
LIQU!O-N2-COCLED
TRAP
· Fig. 4.
11
this period, essentially unattended operation was the rule rather than the
exception. The devices used to stabilize boiling in these systems have been
tailored to specific requirements; but, fri general, it has been desirable
to provide both an independently controlled device to initiate boiling and
a parasitic device, which is actuated by the boiler heat flux, to stabilize
boiling after it has been initiated.
The indep..uently controlled device is normally an appendage ('hot finger")
connected to the system to provide a small, heated volume of liquid metal
that can be taken to temperature levels sufficient to cause generation of
bulk vapor within the appendage. Thus, superheating and flashing occur in
OCUL
a low-volume region wherein energy storage is small and energy release during
flashing is insufficient to perturb the nain system. After a vapor pocket is
established, a stable liquid-vapor interface is available at which vaporiza-
tion can take place. Growth of the vapor pocket results in the release of
vapor into the main boiler pool to prevent any tendency toward superheating.
It has been demonstrated qualitatively that, after boiling is initiated, the
tenperature of the "hot finger" can be brought within a few degrees of the
main pool temperature without flood ing of the device and loss of effective-
ness. Figure 5 illustrates the basic design of an effective "hot finger."
Such a device can be independently heated either by a ceramic-resistance wire
heater mounted directly on the rod or, in a slightly different geometry, by
direct passage of electric current through the body of the "hot finger."
CUCI
Because of the additional control and monitoring problems, it is desirable
to limit the use of the "hot finger" to the startup period and to provide some
other device for sustained operation. Figure 6 illustrates such a device,
a geometric modification of the simple sleeve described previously for use in
ORNL-DWG 65–749

..--
- THERMOCOUPLE WELL
...........
... POTASSIUM --
....... POOL..
•
•
..
-0.052 DIAM * 112 DEEP
LY/2DIAM X 1/2 DEEP

:
"Hot Finger"
Bir
Fig. 5

.
-
-
-
-
HOLES
VIO OIO'O
BOILING RING
(DOUBLE SCALE)
Sincit
darbo d
iena
..
-
-
-
-

Jun 16 (
cartridge heater tests. This sleeve is shrink-fitted to the inner wall of
the potassium boiler in an area of normal heat flux; and, as explained he-
OSS
fore, the tenperature drop across the sleeve is sufficient to maintain a
stable vapor pocket after boiling has been initiated in the system by the
independently controlled "hot, finger". This parasitic device permits long
periods of stable operation with the "hot finger" deenergized, but it raust
be reactivater by use of the "hot finger" if the system is shut down and the
vain dröket becomes flooded.
EFFECT OF ARGON GAS ON PUIT CAVITATION
Because of the relationship of the superheating phenomenon to the safety
of liquid metal fast breeder reactors, it is important to examine the effects
of argon gas in suspension or in solution on the tendency of liquid metals
to superheat. Work is planned in this area at the Argoune National Laboratory
(Ref. 4) and at the Atomic Energy Research Establishment (AERE) at Harwell.
Some related data were produced during the course of potassium pump cests
(Ref. 5) at the Oak Ridge tational Laboratory.
The cavitation characteristics of three boiler feed pumps for use in boil-
ing potassiun materials compatibility test facilities were recently determined
S were ]
(Ref. 5) in a liquid potassium pump calibration loop (Fig. 7).
Two different methods were used to vary the net positive suction head (XPSH)
at the pump inlet.
In the argon pressure method, the suction pressure was
varicd at constant system temperature by changing the argon cover gas pre:s-
sure in the surge (metering) tank. In the vapor pressure methods, the argon
SUIT
was first removed from the system by evacuation through a potassium vapor
ORNL-DWG-66-5102
· TO VACUUM
OR ARGON
.
HOLD UP TANK
.
COLD
TRAP
METERING
TANK
-LEVEL INDICATOR
-OT
EAP
TESTLUS
KFILTERS
15

COCK
PA
- AIR
12"
YVE
.
ORIFICES_*
: PUMP
--SUMP TANK
FLOWMETER
SIMPLIFIED SCHEMATIC OF
ELECTROMAGNETIC PUMP TEST L00?
FICG?
16
condenser-freezer; during the test, the surge tank temperature was held con-
stant, while the potassium temperature entering the pump surtion was varied
to change its vapor pressure and the NPSH.
Figure 8 illustrates two typical cavitation inception curves for the same
44
pump at essentially identical temperature and flow conditions. The results are
affected markedly by the technique used to lower NPSH. With the argon pressure
method, cavitation inception can be assumed to occur at zero angl ii
perimental accuracy. With the vapor pressure method, Operation was possible
at significantly negative values of NPSH. This condition corresponds to the
presence of superheated liquid at the purap suction. For the case illustrated,
200°F at 1300°F..
The technique used to perform the test in the presence of argon cover gas
required that an isothermal condition be maintained from the flow-through surge
tarik to the pump suction. This procedure assured that any argon in solution in
the surge tank would not be released to form discrete gas bubbles in the pump
succion stream. The geometry of the test system required introduction of the
potassium into the flow-through surge tank from the top in a manner that makes
it impossible to discount gas bubble entrainment.
Wall
Based on the test conditions that pertain, the maximum bubble size that
could have been entrained was approximately one mil in diameter. However, a
full-flow filter having a nominal pore size .ange of 10–20 microns was in-
no
stalled in the line between the surge tank and the pump suction. These facts
plus the transient nature of the test and the lack of specific instrumentation
ORNL-DWG 67-5410

STRA
RENJAWAN NUR GEGU vra nje arm ABLECH A Kararan I LINDURHAM MAH.Lemagrue mean
JKK'NILNA
tecomminnes
s temmerne nye
FLOW RATE
43 lb/hr
HELICAL INDUCTION PUMP PB1
INLET TEMPERATURE STATOR VOLTAGE
1300°F
250
.
--Propu orang non so pero
ANALA A A
Orros
3XZ
brzo
.. PRESSURE RISE (psi)
O VAPOR PRESSURE METHOD
ARGON PRESSURE METHOD
1.
I
10
NPSH (psia)
COMPARISON OF CAVITATION INCEPTION USING
*TWO METHODS OF REDUCING NPSH
FI68
18
make it impossible to determine whether bubbles having the critical radius
(Ref. 6) for nucleation from an inert gas-vapor mixture were present at the
pump inlet.
Qualitatively, these results indicate that the presence of inert cover
gas in the pump test facility was an effective deterrent to superheating
owing to its presence in suspension or in solution in the pump suction
stream. This observation is pertinent to the safety of liquidmeta)--anırd
fast reactors since it indicates that the presence of inert gas in the reactor
primary coolant system may cause sodium to boil readily during off-design
power transients rather than to superheat.
Further tests to verify and to define the effects of inert gas on
liquid ñetal superheating are in order.
ACKNOWLEDGEMENT
The test programs that produced the techniques, observations, and measure-
Metals and Ceramics Division of the Oak Ridge National Laboratory. Foremost
among these were J. C. Amos, * C. W. Cunningham, D. L. Clark, L. C. Fuller,
A. G. Grindell, E. E. Hoffman, W. R. Huntley, and R. N. Lyon,
Now with General Electric Company, Evendale, Ohio.
19
REFERENCES
1.
A. I. Krakoviak, Superheat Requirements with Boiling Liquid Metals,
Proceedings of 1963 lligh-Temperature Liquid-:etal lleat Transfer
Technology Neeting, USAEC Report ORML-3605, Vol. 1, p. 310,
November, 1964, Oak Ridge hational Laboratory.
2.
R. E. Holtz, The Prediction of Liquid Superheats Required for Initiation
of Nucleate Boiling in the Liquid letals, paper 64-TA/HT-31 presented
at *e "inter Annual Meeting of the American Society of Mechanical
Engine.rs, New York, New York, November 29-December 4, 1964.
3.
J, A, Edwards and 11. W. Hoffman, Superheat with Boiling Alkali Metals,
Proceedings of the Conference on Application of High Temperature
Instrumentation to Liquid Metal Experiments, USAEC Report ANL-7100,
p. 515, September 28-29, 1965, Argonne National Laboratory.
4.
Argonne National Laboratory, Reactor Development Program Progress Report,
October 1966, USAEC Report ANL-7267.
5.
W. R. Huntley, H. C. Young, and A., G. Grindell, The Cavitation Character-
istics of Two Types of Electromagnetic Pumps in Potassium, Cavitation
Forum, p. 15, paper presented at hinter Annual Meeting of the American
Society of Mechanical Engineers, New York, New York, November 30,
1966.
6.
L. S. Tong, Boiling Heat Transfer and Two-Phase Flow, John Wiley & Sons,
1965.
AT:
#C
.!
..
9 / 28 /167
DATE FILMED
...
..1
.
.'
LK -
ONE
D
1
"-
went
on to.
.
.
- ide on
the imagen de la quimiento en