•
:
| OF I.
ORNL P
1501

.
**
.
-
$
i
"
...
-
PFEFEEEE
L
.
1.4 1.6
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -1963
r
ہ کر۔ ا ا Enه
conf-650910-7
SEP 1 6 1965
HFIR PREOPERATIONAL VIBRATION AND HYDRAULIC
TEST PROGRAM
- LEGAL NOTICE
The report wu popurid un scount of contrucat sponsored mort. Molther the United
main, nor the Commission, nor w pornon sottes au bowall of the Coundaria:
A. Makes my v ruty or representation, 'exprosud or replied, with respect to the socv-
rey, completenes, or wetlands of the dafarutoa catalud de suport, or that the ce
of wy taformation, appunto, method, or procau duolowd la tela report aby bot latring
petrutaly oned rights or
8. Awww way babiliuse with respect to be ww of, or lor dansuring the trou the
w o Worluation, appundus, method, or proceu deslord la de report.
Aoud too the above, porno acting on all of the Coualaston" boludne ay na-
ployee or contractor of the Counastou, or sployw al mich contrrotor, to the adot was
mod eployee or contractor of the Countudo, or employs of much costructor proporno,
w orlas, ar produs soos
or parties now lo, w tloruation purnust to H, waplogant or contract
vtu dhe Commission, or No onploynt with much contractor.

G. J. Dixon
July 28 and 29, 1965
RELBASED FOR ANNOUIICEMENT
IN NUCLEAR SCIENCE AESTRACI'S
Oak Ridge National Laboratory
Operated by
Union Carbide Corporation
for the
U. S. Atomic Energy Commission
HFIR PREOPERATIONAL VIBRATION AND HYDRAULIC
TEST PROGRAM*
G. J. Dixon
Oak Ridge National Laboratory
Oak Ridge, Tennessee
The primary purpose of the High Flux Isotope Reactor. 13 to pro-
duce transplutonium elements (americium, curium, californium, berkelium).
A second and much more general objective is to provide a large number
of high flux type experimental facilities in the reactor without inter-
fering with the production of transplutonium elements.
The HFIR is a light water moderated and cooled, beryllium reflected.'
flux-trap type reactor, which utilizes highly enriched uranium-235 as
the fuel.
The design power level 18 100 Mw which produces an unper-
turbed neutron flux of the order of 5 x 10
neutrons/cmsec in the
island region. The reactor core consists of a series of concentric
annular regions about two feet high.
(Slide No. 1)
A 5-in. diameter
hole forms the core center; or island, which receives the target con-
taining the 444Pu. The fuel region, composed of two concentric fuel
elements, surrounds the target and is encircled by a ring of beryllium.
Two control plates in the form of concentric cylinders driven in oppo-
site directions are located in the annular space between the beryllium
reflector and the outer fuel element.
The reactor assembly is contained in an 8-ft diameter pressure
vessel. (slide No. 2) Primary coolant (light water) enters the
vessel through twin inlets near the top, at a rate of about 16,000
*Research sponsored by the U. S. Atomic Energy Commission under
contract with the Union Carbide Corporation.
8pm, passes through the core and exits at the bottom of the vessel.
The vessel outlet 18 a single 18-in. line which continues through the
tunnel to supply the four individually compartmented heat exchanger-
pump cells. A 10-in. 'line runs from the main header to the heat
exchanger inlet and a 14-in. line runs from the heat exchanger outlet
to the pump suction. A 10-in. line then connects the pump discharge
to the 20-in. vessel inlet header in the tunnel.
The system 18
designed to operate at a pressure of 1,000 psi with normal operation
requiring a pressure of about 600 p81.
Considerable time and effort have been expended in precperational
tests to assure the mechanical integrity of the core assembly and the
entire primary coolant system and to determine in detail the flow
characteristics through the various regions.
Vibration Studies
Prior to installation of the reactor components in the pressure
vessel, the primary coolant system was operated with an orifice
mounted in the vessel to simulate the pressure drop anticipated in
the core assembly. While the orifice was in the vessel, extensive
measurements were made on the primary system piping and the vessel to
determine the location of any serious vibrations. Piezoelectric type
vibration sensors were employed for these measurements so that the
ainplitude and frequency of the vibrations could be determined. Sev-
eral hundred measurements were made during the program.
In most
instances, the sensors used in measuring the system piping were
mounted to detect three planes of vibration. Vibration magnitudes of
up to 9 mils were found in the piping system and up to 13 mils were
found on the vessel. Frequency data indicated that the significant
of pressure fluctuation in the vessel exit line.
During the initial operation, thimbles inserted in the primary
piping for temperature detection and safety system testing broke off
almost immediately. A study of the thimbles indicated that they all
had natural frequencies very close to that of the pumps (150 cps).
When the vessel orifice was removed,' vibration measurements were
again made on the system. The system pressure drop was taken across
the pump discharge valves. The amplitude of the vessel vibration
decreased to 1 to 3 mils and that for the piping decreased to 1.5 to
2 mils. Filling the pool with water decreased the vessel vibration
to less than 1 mil.
While the system was down for installation of the core assembly,
the thimbles' were strengthened to raise their natural frequency to
between 700 and 1,300 cps.
The piping supports were also stiffened
to increase their natural frequency. After assembly of all reactor
internals the piping and vessel vibration measurements were repeated.
Maximum vessel amplitudes, without water in the pool, were 0.2 to 0.7
c o.-.-.-.
mils on the top head. The bottom head amplitudes were 0.4 to 0.5
Windows medi
mils. Piping vibrations remained at 1 to 2 mils and the thimbles were
apparently satisfactory, since they remained intact.
*
Making the external vibration measurements posed no problems ;
however, the design group had asked for vibration measurements on
various parts of the core assembly and this did present some problems.
F
This meant that the instruments and leads had to be made waterproof,
withstand pressure, and had to be mounted and the leads brought out
of the vessel in a way that would offer the least disturbance to the
normal flow conditions. Quantitative vibration data were required
on the vessel extension (lower) shield plug, (slide No. 3) the fuel
grid support pedestal, the movement of the removable reflector with
respect to the permanent beryllium, the deflection of the outer control
plates while being operated up and down, the outer shroud, the target
tower, the flow distributors mounted over the vessel water inlets and
the core support pedestal.
Two types of instruments were utilized
for these measurements--differential transformers and Accelerometers
(piezoelectric type).
The differential transformers were utilized
on the bottom of the shield plug to read the movement of the plug
with respect to the vessel wall through a Delrin "finger" (attached to
the movable core) which was springloaded against the vessel wall.
(Slide No. 4)
The movement of the semipermanent beryllium with respect
to the permanent beryllium was also measured with the differential
transformer. Here the transformer was actually mounted on the reflec-
tor support pedestal with the Delrin fingery springloaded against
the removable reflector. In both the shield plug and the reflector
measurements, two differential transformers, mounted 90° part, were
utilized. A differential transformer was mounted on each wing of the
flow distributor with the Delrin fingers being springloaded against
the vessel wall. To measure the deflection of the 65-in. long outer
control plates, differential transformers were mounted on the lower
ring of the upper track assembly with the Delrin fingers spring loaded
against the lateral midpoints of the control plates.
The differential transformers were all calibrated and checked
under water prior to installation in the reactor.
They were posi-
tioned so that the Delrin fingers could move in or out about 1/8 of
an inch. The leads were brought out of the vessel, via various routes
within the core assembly, through aluminum tubing. Data were taken
-
-
-
at one-, two-, three-, and four-pump operation. The system pressure
was limited to 300 psig and the temperature to 100°F to protect the
vibration instruments from possible damage.
Measurements on the shield plug showed the maximum movement to
be approximately $3 mils and movement of the removable reflector was:
practically nil. The outer control plates were first moved up and
down with no flow to establish the deflection pattern due to travel
up and down the tracks. The measurements were then made at one-,
two-, three-, and four-pump operation. At normal three-pump operation
the control plates were deflecting outward (away from the fuel element)
almost 1/8 of an inch. This was contrary to the design calculations
and predictions. It had been postulated that the deflection due to
hydraulic forces would be inward toward the fuel element and that the
magnitude would be about 40 mils.
Repeated operation of the control plates reproduced the original
data. Since the reliability of the measurements was questioned, it
was decided to go into the reactor and slip a feeler-gauge between
the control plate and tip of the Delrin finger. This action not only
.
verified the indicated direction of plate rovement but also verified
---..
-...
-----
the measured magnitude of the deflection.
An analysis of the orificing arrangement permitting water to enter
the control plate region revealed that a jet effect was taking place,
forcing a high velocity of water between the control plate and the
fuel element, causing an outward deflection in the plates. The orifice
was redesigned to permit a more uniform flow through that region.
.
.
-
Subsequent tests showed that the control plate then moved in the
proper direction and the magnitude of deflection was only about 10
mils.
....ationen innan
Vibrations in the fuel grid support pedestal, the reflector con-
..
.'
tair.er, and the target tower were measured with the Accelerometer.
.
.
. .
These instruments were rigidly mounted in waterproof containers.
The
.
. .
.
.
...
.
.
.
..
exterior surface of the side of which the Accelerometer was mounted
was contoured to fit the outside diameter of the particular piece
to be measured. The boxes containing the Accelerometers were banded
to the equipment being tested and the leads from the instruments
carried out of the vessel in aluminum tubing.
As in the case of the differential transformers, measurements
......
..
......
.
..
.. ..
always utilized two Accelerometers mounted 90° apart.
.
.
..
..
Both the target tower and the outer shroud exhibited vibrations.
.
Correction of the orificing arrangement into the control region (pre-
.....
...
viously ment ioned) elim inated the shroud vibrations.
The target tower
.
vibrations were reduced from 30-40 mils to about 3 mils by using jack-
screws to firmly anchor the shroud spider.
Flow Measurements
While the orifice simulating the core AP was mounted in the vessel,
the flow characteristics of the primary system were also determined.
Fast response sensors were utilized, in addition to the normal instru-
mentation, to measure certain transient pressures and flows. These
instruments, calibrated to accuracies of £2%, were read out on a multi-
channel Visicorder. System flows and pressure drops were measured for
all combinations of the four main circulating pumps for both main motor
and pony motor operation, respectively at a system pressure of 600 psi
and a temperature of 100°F.
The design flow is about 16,000 gpm at
a total system pressure drop of approximately 155 ps1 at a 1988e1
inlet and outlet temperature of 120°F and 169°F, respectively, at an
operating pressure of 600 psi. The four pumps produced almost identi-
cal flows; two pumps gave flows of 9,600 gpm and the other two 9,700
3pm (AP 96 p81). The average flow for two-pump operation was found
to be 14,185 spm (AP = 131 psi), for three-pump operation 15,875 gpm
(AP = 152 psi), and for four-pump operation 16,700 gpm (AP = off scale).
These flows agree reasonably well with those predicted from the vendor's
curves.
Tests were made to determine the point at which cavitation occurs.
As was expected, cavitation occurs at the vessel orifice prior to
occurring in the pumps. The system pressure was lowered from 955 psi
in 100 psi increments until cavitation occurred. The first indication
came at a pressure of 280 psi and cavitation was quite evident at 245
psi. The reason for starting at such a high pressure was to obtain
flow data at various pressure levels in conjunction with the cavitation
runs. These were initially made at 100°F and then repeated at a
temperature of 180°F. As was expected, the higher temperature elevated
the point at which cavitation occurred. At the higher temperature the
first indication appeared at a pressure of 298 psi compared to 280 281
at 100°F.
-
-
The effect of temperature on system flows and pressure drops was
investigated over the range 100°F to 190°F at system pressure of 600
and 900 psig and no unusual effects were observed. The primary coolant
was heated by continued operation of the main circulating pumps. The
pumps are driven by 600 hp motors and by operating all four pumps the
system temperature could be rained at a rate of about 30°F per hour.
The volume of the high pressure system was determined by metering
the quantity of water required to f111 the system. With all four pumps
and heat exchanger cells valved into the system, the volume was deter-
mined to be 13,810 gallons. The "spring constant" or quantity of water
required to pressure the system from atmospheric to 1,000 psi was also
determined.' The actual measurement was made by pressuring the system
to 1,000 psig and bleeding off small volumes of water and noting the
..
. .
.
amount removed along with the system pressure and temperature. About
50 gallons of water are required to pressure the system to 1,000 psig
.
. .
.
at a temperature of 92°F.
.
.
. .
.
The effects of electrical power outages on system flows and pres-
sures were investigated and the coastdown times determined for the
. .
.
.
pumps. This was accomplished by simultaneously shutting off all pumps
.
.
-
and observing the transient flow and pressure conditions on the Visi-
"
...
.
corder. By utilizing time-lines on the Visicorder, coastdown times
..
..
.
were determined for the main circulating pumps. Coastdown times of
....
1.90 to 1.95 are required for the pumps from full-flow to half-flow
and 8.95 seconds are required for coastdown to pony motor flow.
A sudden break in the primary system was simulated by actuating
a remotely-operated, one-inch quick-opening valve mounted on top of
the vessel. The valve was allowed to discharge into the reactor pool.
The test was conducted at system pressures of 600 to 900 psig with the
letdown system isolated from the high pressure system. The pressurizer
pump was left running to ascertain the point at which the system pres-
sure would stabilize. When the valve was actuated at 600 psig, the
pressure dropped to 415 p818 before stabilizing. At this condition
the valve was discharging water at the rate of 300 gpm. When actuated
at 900 psig, the pressure dropped to 600 p81% before leveling and the
flow through the valve was 365 gpm.
The flow rates through the valve
were determined from the flow to the pressurizer pump.
There was no
indication that a sudden release of water at the rates measured would
produce a shock wave in the system. This test also served as a good
demonstration for the operators regarding what might be expected if a
sudden break occurred in the high pressure system.
The pressurizer pumps were exercised through all modes of opera-
tion and the capabilities of the letdown system and the pressure
relief valves determined. The pressurizer pumps performed satisfac-
torily as did the letdown system but the pressure relief valves galled
during their test. They were later modified and reinstalled but galled
again.
The valves would deliver the prescribed amount of water at the
proper pressure setting but galled during closure.
The core assembly was provided with several permanent pressure
taps and some temporary taps for testing purposes. After installation
of the reactor components in the vessel the flow studies previously
made were repeated with additional measureraents involving the various
flow regions in the core assembly. Prior to establishing flow, the
.
.
.
.
.
.
.
system pressure was raised to 900 psi in increments of 100 psi and all
.
. .
-,.
..-.
instruments compared against each other. Although the instruments
. were calibrated prior to testing, the previous test series pointed
out the desirability of knowing exactly how one instrument compared
with another at the same pressure.
The flow rates through the various core regions were about as
expected with the exception of those through the fuel element, which
were lower. A fuel element flow of 12,950 gpm was anticipated and
the actual flow was about 11,750 gpm. The flow through the fuel element
inawet
-
-
-
---
was actually determined by difference.
The flows through all the
.
t
h
core regions except the fuel element were measured and the differ-
ence between these and the total flow through the vessel was assumed
to be that passing through the fuel element. Investigations are now
underway to determine the cause of the low fuel element flow.
The
..... --- mamonavite wine toimited
predicted flow through the control region was 1,800 gpm and that
measured was 1,815 gpm; for the target flow, 788 gpm was expected
and 805 was measured. The predicted fuel element labyrinth flow
(annulus between the two elements) was 172 gpm and that measured was
156 gpm. Where three-pump operation had previously produced a total
flow of 15,800 gpm, the flow now is about 14,800 gpm.
The pressure drops through the core were in general higher than
anticipated, particularly through the fuel element. A AP of 106 psi
was expected through the fuel element and 127 psi was measured.
Another anomaly was the pressure drop across the outer control
rod connector. A pressure difference of 1 to 2 psi 18 required for
heat removal from a stainless steel ball used in this connector. The
measurements indicated that this AP was in the reverse direction.
All flow and pressure drop measureinents were made for one-, two-,
three-, and four-pump operation. Studies for three-pump operation
were made ar: 99, 120, 152, and 198°F. The cavitation studies were
repeated at temperatures of 100°F, 150°F, and 180°F. These tests
were terminated at a vessel inlet pressure of 158 psi because of
possible damage to the pump seals. At this point the pump suction
pressure was essentially zero and there was slight evidence of pump
cavitation.
.
,-
' '*
Conclusions
er via
It is felt that the test program, thus far, has accomplished
its purpose. Several areas where improper flow characteristics or
mechanical instabilities due to flow existed have been found and the
necessary corrections made. Further testing will be done in regards
to the fuel region flows and some other minor anomalies that still
exist.
The test program also provided the operators with the opportunity
to become faimilar with the equipment and to get the "feel" of various
tools used in the remote handling of the core assembly. All operating
crews performed the refueling operation several times during the
program, as well as removal and installation of other core components.
- -
-
en to me
t
Lis
intir
ORNL-LR-OWG-65776
UNCLASSIFIED

TARGET RODS
OUTER CONTROL ROD
INNER FUEL PLÁTES
PENYOUTER LES
-OUTER FUEL
111
TILDI22
-
-
BERYLLIUM
REFLECTOR-
-
.
PANI
_
-------
--
-
-
-
---::::
-.-.
BEAM TUB
.-
-
.
...
..
INNER CONTROL ROD
slide No. 1. Schematic of High Flux Isotope Reactor core.

O
.
mone
0:1
T
UNCLASSIFIED
ORNL·LRDWG-65339
•
...
CIRCULATING
PUMP (TYP.)
HEAT EXCHANGER
CELL (POR
FUTURE USE)
FLOW VENTURI
HORIZONTAL BEAM
CAVITY (TYPICAL)-
ننننننفسشنید
Slide No. 2.
THEAT EXCHANGERI.
SECONDARY
WATER TO
COOLING
TOWERS
.
..
..
....
TOOL STORAGE
000
1900
SPENT FUEL STORAGE
OOOOOO
olololololo
REACTOR PRESSURE
vesse
CLEAN
CRITICAL
FACILITY
REACTOR
POOL
CALE: (FEET
HIGH FLUX ISOTOPE REACTOR
RFACTOR SHIELD, HEAT EXCHANGER CELLS, AND POOL
STRUCTURES (HORIZONTAL SECTICN)
QAK RIOGE NATIONAL LARORATORY
-
.
.
.
.
.
.
.
.
..
.
.
.
...
.
.....

LREACTOR POOL
LL ,...
QUICK ACCESS
HATCH
FUEL
STORAGE
WATER
INLET
FUELLT
IN ELEMENT3 Klarneta
CONTROL
TARGET FL
Auto
ROOS
HOT
mo
NON CHAMBER
THIMBLE
.
BERYLLIUM
REFLECTOR
T PRESSURE
VESSEL
OUTER CONTROL
ROD DRIVE
.
WATER
OUTLET
INNER CONTROL
ROD DRIVE ~
.
11
CONCRETE
SHIELDING
MVT
FISSIONS
CHAMBERS
THIMBLES
1.17..de.
HFIR
ASSEMBLY
PROBE A
Talli
.
DUTETT
Slide No. 3 :
..
.....
.
t
:.:::..
.........
.
. . momsormia voi.i
ww.......
,
P
.
..
wa
intimita's wet w
***

+
Fig.az
...
OAK RIDGE NATIONAL LABORATORY
...
:
.
...........
3.
Y
.
1
END

.
.
.
......
-wo.........
DATE FILMED
10/25 / 65
:
•
i
.
...
L
STAND WI