:ii
:
| OF T
ORNL P
2978
.
.
:
EEEEEEEE
11:25 .1.4 ILLE
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -1963
- ORNU-Pa 2978
CONT-670501-06

MASTER
APR 2 8 1967

UTILIZATION OF THE CAK RICE RESEARCH REACTOR TO STUDY T:
MECHANISM OF FISSION GAS RELEASE FR
V PIRATON RELEASE TROV REACTOR IT318
1
N
CESTI PRICES
R. M. Carroll R. B. Perez'
0. 8: aran
G. M. Watson
Oak Ridge National Laboratory
Oak Ridge, Tennessee
HC 63.00: MN65
INTRODUCTION
The release of volatile Pission products from the fuel is one of the
important factors which determine the operating life oť metal-clad ceramic
fuel elements. If all the noble fission gases escape from the fuel struc-
ture they can exert considerable pressure on the fuel cladding. For example,
this pressure would be several hundred atmospheres after 10,000 MWD /metric
ton at 1000°C, if the fission 888 18 confined to A volume equal to 5% of the
fuel volume On the other hand, if unclad or vented ceramic fuels are.
used, as proposed for modern high-temperature gas-cooled reactors, the re-
leased volatile fission products may escape from the fuel region and con-
taminate the primary coolant system. This contamination is in the form of
ga ses, which increase the hazards by rupture of the coolant circuit; also
zon-volatile radioactive decay products, which deposit on the coolant cir-
cuit surfaces, will prevent or hinder maintenance and increase shielding re- ,
quirements.
. An experimental program is in progress at the Oak Ridge National Lab-
oratory to evaluate the parameters that control the release of fisslon gas
during irradiation in the hope of developing a model to predict fission-gas
release from ceramic fuel elements. In a parallel program complementary
studies are being performed on pyrolytic-carbon-coated fuel particles to :
evaluate the integrity of the various types nf pyrocarbon coatings by study-
ing the fission-gas release during irradiation.
EXPERIMENTAL METHODS
Fission-gas release is measured during irradiation of the fuel with
close control of fission density and temperature. 12 The fission density is
regulated by moving the specimen (UO2 or coated particles) to the appropri..
ate position in the neutron flux of the Oak Ridge Research Reactor. The
specimen is heated by its own f18bion heat and the temperature is controlled
by air-cooling of a capsule which contains the fuel under investigation. A
continuous flow of sweep gas carries the fiosion gas outside the reactor,
where it 18 analyzed by gamma-ray spectrometry. Figure I shows & flow dia-
gram of the apparatus and Fig. 2 shows & cross section of a capsule configu-
ration which is used for 102 fuel.
"Research sponsored by the U.S. Atomic Energy Commission under contract.
with the Union Carbide Corporation.
b onor a momentas UNAM
· Consultant, University of Florida, Gainesville, Florida.
-
--
-
-
-
LEGAL NOTICE
...
ORTL-DWG 64-7276
Tudo report w prepared u an account of Government sponsored work. Nedohor the United
rates, we the Counluston, nor any person acting at baball of the Counteeton:
A. Makes my narranty or representation, expressed or implied, wla rospect to the accu-
racy, completeness, or watalnost of the information contained ha the report, or that the we
of any information, apparatu, method, or procesu disclosed in the report may not tortugal
privately owned this; or
B. Asmuna ay labtutiar with respect to the wo of, or for domes relating trou the
an al my twormation apparatus, nethad, or process inclound in the reporte
As used to the above, person nethegou brokalt of the Commission" moked my
plom or contractor of the Contacto, or Wsployee of me cominctor, to do that
much aployee or contractor of the Co nstan, or neploys of mucha contruotor propera,
dorbate, or provides woww to, any traformation par to Mo amployment or contract
with the Commtendon, or Mo saployment with much contractor.
HYDRAULIC
POSITIONER
PROCESS WATER
(REVERSIBLE FLOW)
HIGH-LEVEL
COUNTER
PURIFICATION
SYSTEM
CHARCOAL
TRAP
ю
STACK
I.
ARGON
HELIUM
GAMMA
SPECTROMETER
.
.
AIR :
SUPPLY
COUNTER
1

COOLING
AIR OUT......
Το
STACK
UO, SPECIMEN-
FLOW DIAGRAM OF FISSION-GAS RELEASE EXPERIMENT
COM
Frei ?
.
. . .

UNCLASSIFIED
ORNL-LA-OWO 40238R3
SWEEP GAS
IN
THERMOCOUPLE LEADS

AIR
.
.
.
S
S
AIR OUT
wwwwwww
MIT
1
TUTT
_SWEEP GAS OUT
-POSITIONING TUBE
II
:
SC
We
M
VW
-SWEEP GAS OUT
Al2O3 HOLDER
-SPECIMEN CONTAINER
2
MINIMINIINTİMwin
INSULATED THERMOCOUPL
STIL
REACTOR.
.-
CENTER LINE
-
PECIMENS
DI
1
CROBS SECTION OF FISHICH-GAS RELEASE CAPSULE
.
The neutron flux at the specimen is measured by using & sweep cas vith
a lmoun argon concentration and by measuring the neutrop activation of the
argon, (The argon peak my be recorded along, with the fission-gas spectru
and is, therefore, a very simple measurement. )
The VO2 specimens are selected with great c
since it has been
well established that the physical properties of voz (e.g., density, surface
area, and oxygen-to-uranium ratio) have a very strong effect upon fission-
gas release. To reduce the number of parameters affecting the gas releane,
all the specimens are of nearly theoretical density (10.96 g/cm® ), have an
0:V ratio of less than 2.01, cortain no detectable porosity, and the total
impurities are less than 200 ppm. The specimens are machined into thin .
discs 0.10-cm thick and 1.27-cm diameter.
1
.
The pyrolytic-carbon-coated rei partic..es studied are divided into
three, types according to coating construction: monolayer, duplex, and tri-
plex. 107 Some coatings may have a buffer zone of porous carbon between the
fuel core and the initial pyrolytic carbon coating. Coa
also classified according to the features of the pyrolytic carbon coating.
Laminar coatings (sometimes called onion skin) may be either isotropic or
anisoiropic. Columar coatings are normally anisotropic. Granular coatings
are also anisotropic but show less columnar features.
.
The experimental method for the coated particle studies has been de- .
scribed previously in detail. 5,6] About 2000 parti
graphite holder inside a steel capsule, see Fig. 3. The capsule 18 inserted
in & lattice position of the Oak Ridge Research Reactor (ORR). The tempera- ::
ture of the fuel particles 1s monitored by five thermocouples located in the
graphite holder; one of the thermocouples is surrounded by the fuel par-
ticles and is considered to be at the maximum fuel temperature.
EXPERIMENTAL RESULTS
Fission-gas release from UO2
:
•
Early investigations of the high-temperature release of fission gas
: from U02 showed that the gas was emitted as an exponential hection of tem
perature, and it was generally accepted that diffusion was the primary fis-
Bion gas release mechanism. The results from in-pile experiments led to the
development of the "Defect-Trap Model” in place of the diffusion model.
These resulty, which have been described in detail in previous publica..
tions, •,7 J'will be outlined below,
..
During irradiation, there are two basic processes by which fission gas
18 released from VO2, Bee Mg. 4. By classical theory, these two processes:
would be described as recoil release (temperature independent) and dirrusion
release (temperature dependent). The
portional to nosion nte, and this process is responsible for the njor por-
tion of the gran release at temperatures below about 600°C.
.
CIRNL-LR-DWG 57280R
HOLE FOR VENTILATION
OF GAS (2) -
- HOLES FOR
THERMOCOUPLES

:
- LID (GRAPHITE)
-CAN (GRAPHITE)
-
-
-
-
- SLEEVE
(GRAPHITE)
........
kunne s
CENTER PIN
(GRAPHITE)
INAH...
FUEL, SPHERICAL UCZ
COATED WITH PYROLYTICALLY
DEPOSITED GRAPHITE
Coated Particle Capsule for ORR C-1 Facility: Type 2.
.
. ORNL-DWG 64-7276
+ O FLUX = 2.5 x 101.1 neutrons/cm2. sec –
• FLUX= 5.0x1013
RELEASE RATE (atoms/sec)
5x105 L
500 600 700 800 900 1000 100 200
TEMPERATURE (°C)
Release Rate of 88 Kr from Fine Grain Specimen (C1-12) of .00.2.

temperature independent release
Mission as escapes from the fuel structure by a recoil proce88, since,
When rission occurs within about 10 u of the surfacO of 102, it is possible
for a fission fragment to recoil free of the hel. The amount of recoil
· Gscape can be calculatod a...lly, but such cal uations are often misleading
for the following reasons: (1) The fission fragment usually leaves the UO,
surface with Buriciont energy to embed in solid surfaces near the specimen
and, thus, will not be released; (2) When the fit 3ion fragment leaves the
VO2 surface, an average of about 2000 102 m.lecules are ejected, and f18sion
products in this lock-out zone are ejected along with the VO2.10)
We can determine if the fiseion gas is liberated by direct recoil or by
knockout by cornparing the relative amounts of different isotopes in the fis-
sion gas. This is because the different isotopes have different radioactive
decay rates. A direct recoil process will liberate the isotopes 88 suon as ..
they are formed, and therefore, the amounts of the different isotopes in the
11881on as will be in the retio of their fission yields. If there is some
delay time before the gas escapes, then the 1sotopes with shorter ball-lives
will decay, leaving a higher ratio of longer-lived sotopes. The compari.
son shows that, at UC, temperatures below 600°C, most of the P188ion gas is .
emitted by the lock-out proces8.19]
The amount of gas liberated by lockout is directly proportional to the
total surface area of the fuel, whereas the amount of gas liberated by di-
rect recoil is proportional to the geometric surface. We find that the tem
perature-independent release, although directly proportional to the fission
mate (Fig. 5), is not proportional to the geometric surface. This indicates :
that the temperature-independe
mature-independent release is by lockout rather than by di-
rect recoil. The total surface areas of the specimens shown in Fig. 5 are
quite different, even though the geometric surface areas are the same.
:
temperature-dependent release
:
tained by subtracting the temperature-independent contribution from the
total measured gas release. An Arrhenius plot of the temperature-dependent.
gas release is shown in Fig. 6.
It was first assumed that diffusion in toz occurred by a vacancy pro-
cess. To test this theory, specimens were irradiated at different fission
rates, but at constant temperature. It was thought that the higher fission
rates would accelerate diffusion because the fissioning process creates va
cancies. Contrary to this theory, we found that a higher fission rate ap-
parently retarded diffusion so that the diffusion coeffiofent appeared to
decrease in direct ratio to the increase of rission rate. What was
actually being observed was a constant gas-release rate, even though the
rate of production of gas was increasing with increasing fissioning rate
(Figs. 4 and 6).
The results of Fig. 6 are directly opposite that expected for the dir-
.fusion of atoms through a matrix. Regardless of the mechanism by which
.
.
11
...:
.
.
.
(x105)
.. UNCLASSIFIED
ORNL-DWG 64-7280
~0
FINE GRAIN
88RELEASE RATE (atoms/sec)“
-
7-SINGLE CRYSTAL
no
.... 1 2 3 4 5 (x1093;
THERMAL. NEUTRON FLUX (neutrons/cm2. sec?
Comparison of 88 Kr Recoil Release Between Single Crystal
and Fine Grain UO, Specimens of the Same Size and Density.

..
Fig.5
.
....
diffusion occurs, the release should be proportional to the concentration of
the gas in the specimen. The data in Fig. 6 show that the concentravion oť
BOX within the specimen was changed by more than & factor of 4 without pro.
ducing a significant change in the release rate. This and related observa-
tians lod to the formulation of the Defect-Trap Model for fissiou-gas re-
lease frou VO, during irradiation.
defect-trap model
The Defect-Trap Model assumes that the rate-determining process in the
release of fissiun gas from VO2 in the temperature-dependent region depends
on the relative probabilities of the gas being trapped by and released from
traps. Three categories of traps are postulated: (1) intrinsic traps, which
do not move at temperatures biolow 1500 °C and which are voids, grain buun-
daries, and other intrinsic efects of the meterial, (2) point defects which
are formed by Pission fragmente, and (3) clusters of point defects. The
point defects and clusters are formed by irradiation (Fig. 7). Accordingly,
the Defect-Trap Model postulates the presence of both permanent traps and
point defects to explain the fact that only a relatively small fraction of
the fission gas is actually released from the sample. The relative inde-
pendence of the release rate on Pission density is accounted for by the 88-
sumption that the concentration of defect traps depends on the balance be-
tween their production and their annealing.
.:
. An atom generated by fission will diffuse through & trapping matrix,
decaying into a daughter which will continue to diffuse. The material bal-
ance equations for the concentrations of mother, daughter, and point defects
are given in general by:!101
.
(1)
ov/dt = Dv + 8 ,
.::, where the transposed vectors **, s* are:
von T - [M(2,t), H(2,t), Nex/t)] ,
5- {[F(t) = 826t )(2,t)] , [4F(t) - 82(t)H(2,t)), oft)} .
::. P is the linear matrix operator
- ( +80) 0 : :
+ D -+)) os
.
Equation (1) in explicit form 18
- Htz(t) M(2,t) + Bxf(t) - 8 M(x,t)
UNCLASSIFIED
ORNL - DWG 64 -7279
THERMAL FLUX (neutrons/cm2. sec)

i 1.9
• 2.4
A 3.6 x1013
0 5.0
~ 6.0
88kr RELEASE RATE (atoms/seci
51,000 cal/mole ACTIVATION
ENERGY
nai
..5
6
7
8
9
10
11
12
10,000/ (OK)
Temperature Dependent Release Rate of 83 Kr from
Fine Grain Uoz.
[- (7 + 8)] =(2,t) - M, (+) (2,t) + 4xVt) + 2/2,8) - #62,t)
a F(t) -
Mer(t) -
Nur(t).
:
...
...
::. The products AF, BF, and an are production terms for the parent,
daughter, and point defects, and terms of the form (D 82/aze) correspond to .
ordinary diffusion processes. The 1088 terms correspond to (a) radioactive
decay terms such as M(2,t), 2H( 2,t); (b) trapping by permanent defects
Bo M(2,t), 80 8( Z,t); and (c) tripping by point defects - h Ny(t) M(z,t),
H(2,t). The production rate of point defects was assumed to he pro-
portional to the fission density. The rate of annealing was assumed to have
the form of an Arrhenius factor involving the frequency of collision vor of
the free atoms with the potential barrier surrounding the point defect.
The solutions that will be discussed apply to a thin plate of material
with spatially uniform temperature and fission density distribution. The
boundary conditions associated with the problem are symmetry of the concen-
tration profiles with respect to the midplane of the sample and vanishing of
· the parent-daughter concentrations at the boundaries of the specimen.
Solutions to the set of equations (1) have been previously published
the resuits predicted by the steady state solutions appears to be deceptive-
ly good, aş 18 shown in F18:,8. However, as has been discussed by Carroll.
et al.123] and by Perez, 1104 an apparently good fit of the experimental.
points with the predicted curve does not constitute an adequate test of the
theory. The reason 18 that there are too many empirically adjustable parane-
ters in the equation which facilitate not only a good fit but a number of
equally good fits.
the oscillation method
:: In order to test the model, a new degree of freedom has to be introduced
80 that for each pair of values of the fission density and temperature, more
information will become available. This was accomplished by the use of an .
oscillating technique which introduces the capability of oscillating the tem-
perature and the fission rate in the specimen and measuring the release rate
at varying oscillation frequencies.
Extensive modifications of the experiment controls were required to
allow a precise oscillation about a selected steady state condition. The
fission rate is Oscillated by moving the specimen in or out of the neutron
flux. Because the movement of the specimen 18 very slight, the change of
flux is proportional to the movement, and a sinusoidal oscillation of the
Specimen produces & sinusoidal oscillation of neutron fit
The specimen temperature 1B oscillated by a controlled cooling air flow..
over the specimen capsule (Fig. 2). The specimen temperature is measured by
---------
.
ORNL-DWG 64-1957A
..
(1) INTRINSIC DEFECTS
CLOSED PORES,
GRAIN BOUNDRIES, ETC.
NOT MOBILE BELOW 1500°C:
CHARACTERISTIC OF
· THE MATERIAL ..
.....(0) GRAIN BOUNDARY.
• (2) POINT DEFECTS:
CREATED BY FISSION
MOBILE AT ALL TEMPERATURES
• TEMP. DEPENDENT DECAY
• (T,F) DYNAMIC EQUILIBRIUM
011
DO
(3) CLUSTERED POINT DEFECTS ...
SPECTRUM OF SIZES
- MOBILITY DEPENDS ON SIZE
LARGE CLUSTERS BEHAVE LIKE (i)
· SMALL CLUSTERS MORE LIKE (2) ..
.(3).
Defect-Trop Characteristics
Fig 7
.
-
.
. .
.
.
:
,,
.

-
.
. -
UNCLASSIFIED
ORNL-DWG 64-1958R
--$=2.0 x 10^3 neutrons/cm2. sec
LINE IS MODEL FIT, POINTS
ARE EXPERIMENTAL DATA
RELEASE RATE (atoms/sec) Kr88
440
480
520
760
800
840

560 600 640 680 720
SPECIMEN TEMPERATURE (°C)
Uog Single Crystal C1-9.
-
.
COMPARISON OF THEORETICAL MODEL WITH EXPERIMENTAL DATA
..
.
.
.
-
---
.
..
.
-
-
- -
-
-
.
-
-
-
-
-
Fig. 8
.

.NU
..
. WITH
Table of Nomenclature
1
M(2,t) - Concentration of parent nuclei (atoms/cm) :
8(2,t) - Concentration of daughter nuclei (atoms/c%)
Nur(t) - Point defects (traps/cm®), (Nro - Ntr at steady state
Diye Da - Diffusion coefficients for M and a (cm • sec-2) - D'exp-(8E/RT)
DE : - Activation energy for diffusion (cal/mole)
Bye By - Fission yields (flssion-fragments/fission)
- Decay constants (sec-2)
Von ... - Time constant for trap annealing = Vo exp - (AE1F/RT) (sec-2)
DET • Activation energy for trap annealing (cal/mole)
:a. Tráps formed/118sion
&.. Trapping probability from intrinsic defects (sec-a) .
b Second order rate constant (for trapping processes) •
(cm®/trap x sec)
Re[] Symbolizes taking the real part of the complex term inside the
bracket
R:
&lt)
- Gas constant (cal/mole *K)
= N2(t) (secº2) - &
- Absolute temperature (°K):
Time (sec)
:...t
R-
T'
RY.
.
thermocouples, which are connected into a feedback system to control the air.
cooling automatically and produce the desired temperature. Thus, the ten-il
perature can be programmed to oscillate about a given value with a selected..
amplitude, frequency, and wave shape. .
The fission-gas release from the specimen during oscillations is mea-
sured by continuous gamma-ray analysis of the flowing sweep gar. The gamma-
'ray spectrometer is programmed to measure the time-dependent variation of a
given isotope of the fissionals (usually BKT). Tae delay time required for
the fission gas to flow out of the reactor is determined by using sweep gas
that contains argon and by measuring the time lapse between position oscil-
lations and argon-activity oscillations. A precise record of time-tempera-
ture or time-flux 18 kept with a digital recording clock-voltmeter. The
fission-gas release rate measurements are thus sychronized with the tempera-
ture of flux oscillations of the specimen.
.
.
|
Solution of the set of equations (1) at constant temperature and oscil-:
lating flux was obtained by the use of first order perturbation theory.
• The solution predicted experimental differences in the amplitude and phase.
shirts of the transfer noction of the fission-gas release as a function of
frequency of oscillation, (748. 9). Bowever, due to experimental difficul.
ties in maintaining constant specimen temperature, these isothermal tests.
bave been unsatisfactory.
Solutions have also been obtained (12) for the set of equations (1) for
the case where both the flux and temperature oscillate. The solutions are
in the form of series expressions in space-time configurations. To obtain
the solutions, the space dependence and the time dependenre were eliminated
and the equations (1) were converted into a set of alegbra.ic equations in
terms of the Fourier coefficients of the concentrations of parent M, daugh.
tėr #, and traps, Ntr Finally the release rates Ay and its were obtained in
torms of the Fourier coefficients of M, 8, Mtr, D, V, F, and T.
na =3* [ups trases amesema hayo yo).
ha - Ę (-2)* [3 coram os LDS * +2 3 Cols are coming
e ls e lementos
hoe om o { (-2)* [ts costas a es el estados os as far as the theo
D(2) met die
ORNL-DWG 65-4165
- DIFFUSION- TRAP THEORY
- DIFFUSION THEORY.

T=1100 °C
y el 200110W
900 °C
1003 (loy
10-4
2
:
5
10-2
5 10-3 2
w (radians x sec")

PHASE SHIFT (deg)
DIFFUSION-TRAP THEORY
DIFFUSION THEORY
-T=1100 °C
ILT=900 °C
104
10-?
2
5 10–3 2
5
w (rodions a sec).
Release Rate Transfer Function vs Frequency
.
Release
i
TE
COATED PARTICLE STUDIES
:
The coated particle studies are more applied in nature, but the experi-
imental equipment is equally well adapted for this work as for the more basic
studier, with 102. It is necessary to control and monitor the neutron flux
and tenperature in the coated particle experiments just as carefully as in
the 10e experiments. The position of the experiment in the reactor is con-
stantly adjusted to maintain a constant fuel burnup rate at the desired
operating temperature. This adjustment corrects for depleting fuel by burn-
up and for the change in Deutron flux during the reactor cycle. These ex-
periments are performed with enriched fuel and fuel burmups of up to
achieved.
1
Pri
Photomicrographs of two types of coated particles are shown in Figs. 10
and 11. The fuel particles are about 200 u in diameter, and the coating is
about 100 u thick in these illustrations. A duplex coating is shown on a
UO2 particle in Fig. 10. This coating is composed of a porous inner layer
and a dense isotropic outer layer. The coated particle in Fig. 11 is
Ce with a triplex coating. This coating also has a porous inner layer. The
porous inner layer gives spece for the accumulation of pission gas and for .
fuel swelling, and acts as a barrier for fission recoil atoms.
During irradiation the fission-gas release is monitored continuously
and samples of gas are collected periodically for analysis of the Kr and Xe
i isotopes. Small spontaneous activity bursts (see Fig. 12) are caused by the
coating rupture of single particles.lº] These bursts contain about 1.2 x.
10°5 curses of gas, whereas, after 30 days irradiation time, one coated par-
ticle contains about 1.1 x 10"' curies of gas. Irradiation of uncoated par-
ticles demonstrated that less than 10% of the generated gas was released
from the uncoated uranium carbide particles. If à particle should store 1
8 produced, beneath the coating and then suddenly have a coating
rupture, this could cause the observed activity burst.
The ratio of the isotopes released from each type of coated particle
determines the slopes of the lines shown in Fig. 13. Although the fraction-
al release rates from the uncoated uranium carbide particles were much great-
er than those from coated particles, the slope of the R/B (Release Rate/
Birth Rate) ve half-life curve for the uncoated material is very nearly
parallel to the curves obtained from both columnar, and duplex coatings. Ex-
cept for the laminar-coated particles, all specimens released fission gas in
almost the same proportions. These proportions would change if different
coatings required different times for the gas to pass through them. This in-
dicates that the escape of fission gas is through breaks in the coatings:
rather than diffusion through intact coatings.
:
When the fractional fission-gas release for the various isotopes 1s
plotted against the half-life on, a 10g scale, we find the points will follow
one of two characteristic lines.15) If the experiment contains particles with
· broken coatings, the points show & nearly straight line, like the one shown
In Fig. 14 for uncoated particles. However, 11 there are no brokon coatings,..
the line will show an offset between the kryptons and the xenons, indicating.
that the kryptons diffuse through the carbon coating faster than the xenons.
Typical examples of this discrimination between the gases is shown by the..

:
.
-
11
6.
QOM INCHES —
18TT20N
..
.
.:..:.
....
.,
.
.
::.:
oooi -
..:::
.
.
F
..
H
....
.
*Note
ardware
ore.....!
..........
-
i
...
n
.
..
.
.
- ...
a018 INCHES
-
:
.:.
militants container
.
.",
mission.com
disco.decembaik... Women
.
Pyrolytic Carbon Coated Uranium Oxide Particles from Batch OR-348
Mag. 200 X.'(a) Voirradiatod (0) Irradiated to 9.4. $ burnup at Asoc.:
110 do powie 9.27. . . .
. . .
. .' ..
."
::

.
iboration. comma
.
d
e
m
Xool Land
S HON! 8:0°C
m
MOO2 bottito
.
EEE$3HNICIEEEEEEEEEEX002 EE
*
1
..,
.
.
.
.
R.SI
.
1
T
.
N
*
*
*
WIN.
.44
..
..
11
S
.
***
17
in: tul:
.
!
21
-
-
:,
1.1im
.
dir
.
į:
.
.
-
..
-
.
WHAT
.
I.
...
.
.
.
NI
.
.
r
YEARS
.
.
.
IT
14.
.
4
.*/
.....
'
....
.
.
"
N
.
t
A
Ni
25
11
PIN
4
.
'
* *
is.lt
is.
.
"
..
.
.
.
.
.
5:53
.
..
..
.
th
?
S
..
CM
Stan
:
.
Y
*
'
S
'
.
.
NEWS
i
.
.
.
i
..
.
*
i
.
s
it
HS
!
.
-id
:2
2
-
1
.
.
--
indd
ger
m
.
--
.
S
...
6
.
S
*
.
..
.
.
W
.
.
...
hein
.
.
.
!
.
.
.
.
*
.
..
.
.
..
.
.
-.
.
.**
.
M
Tv.
.
.
r
.
ris
.
.
.
.
.
.
4
..
..
.
2
.
.
2
.
"
.
.
*
m
.
***
.
.
-
::
-
-
.
.
.
.
.
-
.
*
.
ue
.
i
.
.
.
.
.
.
.
.
.
!
2
..
..
..
1..
.:
..
.
.!
.
.
.
.
.
..
..
.
.
.
.
.
.
i.
.
S
.
..
.
.
*
.
.
47.
:-
.
+
A
.
CIT
.
A
NI
'
.
.
.
US
.
-
.
11
.
?
.
TI
.
NG
.
.
.
.
11
PID
.
Pem
.
.
.
.
!
A
TA
CO
Pyrolytio Carbon Coated Thoriun-Uranium Carbide Partic108' I rum in me
Batch GA-310 . Mag.200 X. (a) Unirradiated (0) Irradiated to 10 %
.
.
.
w
.
.
..
6
U
..
.
-
'it
"
.
.
:'
*
L
7
SY
Net
.
.
"
-
.--.
.
L
.'
.
C
2
2
T
M
.
.
.
.
11
.
.
.
i
S
.
****
.
.
$
.
.
O
12
.
i
.
...
.
.
.
..
SO
-
.
..
.
*.
.
1.
SU
"
1
..
'
UUM
."
.
.
.:.
.
.
S
UI:
.
.
1
11-1d'ornedos ut.D. Stil 10 dnama
bi
44
71
.
.
:
.
.
0
.
1
..
.
15
.....
.
..
i '
L
.
.
*
.
**
.
.
.
.
.
.
.
.
.
.
.
.
.
...
.
.
2
.
*
-
.
*
.
.
.
.
I
.
YS
..
.
*
..
.
.
2
ANN.
1
..
A
.
*
7
.
.
.
---
.
.
.
-
.
.
.
X
*
C
11
...
*
**
f
,
PI
*
'
"
29
.
...
'
.
.
.
.
.
.
.
.
L
.
sh:
.
.
.
.
A
:.:.:.
Hi,
.
1
..
.
.
.......
.
.
.
.
.
.
."
.
*
"
ini,
.
---
90
vis
.
14
.
.
.
....
.
.
R
I
1
.
.
.
..
:
.*
.
.
.
1:
11
mi
1
7.
261ESK
.
.
.
.
?i
.
.
...riorem reve.
.
.**
::
R19263
11. biny
T 12
.
Voit
megtapas.
....
.-.
:
17
T.
*
**.
tu
1
;
'
.
.
:
I
"
.
.
ht
=
.-,.
L
t
*
.
.
-
!
-
-
-
- -
-
.
.
.
.
-
-
---
-
-
-
"
.
-
.
di..
.
..
..
-
1:2
--'.
.
UNCLASSIFIED
ORNL-LR-DWG 67246R2

WWW.
SWEEP HELIUM ACTIVITY (mr/hr)
TEMPERATURE: 1650°F
FLUX: 1.8 x 10'S THERMAL NEUTRONS/cm2. sec
2
3
25
TIME (hr)
Activity Bursts Indicate Failure of Individual Particles.
The bursts of actixity were rich in gases with long half-lives,
indicating the gas had been confined for a long time before
release. .
Fig 12
pe:
....!
.
...
UNCLASSIFIED
"ORNL-LR-OWG 73168A
:10*1

BARE URANIUM CARBIDE 63.5 % BURNUP OF U
MINDI
DINI
.
:,.
.
:.:
F
R/B, RATIO OF RELEASE RATE TO BIRTH RATE
LAMINAR COATING
F 1~12a/ BURNUP OF USE
T CRITERION XX
COLUMNAR COATING (~80/0 BURNUP OF U)
WLAXALAIDV
*HTH
SELECTED_PBRES
LOUPLEX COATING (~13 a/0 BURNUP OF U.)
Kr86m. TTTTTT
!
X135
Kr88 |K, 87
7 x 33
10-6 lul
105
Pag 13! Doamrown ----
104
HALF LIFE (sec)
alon --- ----- --
om Amronmrrin
m
e
from
--
.
....?"
UNCLASSIFIED
ORNL-DWG 64-1096R2
40

BARE URANIUM CARBIDE. (4 a/O BURNUP OF U)
.
..
1100
.
.
.
.
..
R/B, RATIO OF RELEASE RATE TO BIRTH RATE
TRIPLEX COATING
120.6 /0 BURNUP OF U
"LE TIITTI 280 mm
DUPLEX COATING-
(20.8 a/ BURNUP OF U)
H
- 2500 Ft
The
X8133
10-6 W
106
i Kr85m 17
x33LK-881 KẾT
- 104
HALF LIFE (sec)
103
103
Relationship Between R/B and Half Life for Release
of Inert Gasses from Pyrolytic – Carbon - Coated
Uranium Carbide Particles.
lover curves in Fig. 14. The level of uranium contamination in the particle
coatings is such that this source accounts for all of the lisrion-gas re-
lease observed from unbroken coatings.
The metallic fission products are releaB 3d quite freely at elevated
temperatures. Recent experiments have therefore been designed to study the
release of such fission products as Ba, Sr, Cs, ce, Zr, etc, during irradia-
tion.235 The design of this type experiment 18 snown in Fig. 15.
designed to study the
In this experiment the graphite fuel particle holder is iso).ated from a
graphite deposition plate by an annular gas gap. The components of the 88-
sembly, including the fuel particles, are acid leached or dissolved for the
recovery of the fission products after irradiation. This will show whether
the individual.r18sion products will escape from the fuel particle, will
cross the game gap to the graphite deposition plate, will move farther to the
cooler regions of the metallio capsule or cooling coil, or will even be
carried as far as the filter.
ORNL-OWG 66-4069

--FILTER HOUSING
...
viivi
-
-FILTER THIMBLE
2S ALUMINUM
11
- FILTER
FLANDERS AIRPORE
- -
-
-
-
FILTER HOUSING
EXTENSION
- COOLING COIL, 12 TURNS
STAINLESS STEEL TUBE
. 3.2 mm OD
-
-
- -
CAPSULE CAP
INCONEL
- STAINLESS STEEL TUBE
THERMOCOUPLE HOUSING
AND HO INLET ..
-
-
-
-
-
..
--- .
INCONEL CAPSULE
-PyC SHEETS
B
THERMOCOUPLE SHEATH
RHENIUM FOIL --
-
-
OXIDE WOOL-
-HOLDER CAP
-
--
to
--
-
.
.
.
COATED PARTICLES-
.
Ill..I lol.1.1.1.1.11
OOI.1.
ULL
LATURIT
DEPOSITION PLATE
THERMOCOUPLES (2).
CENTER
THERMOCOUPLES (3)-
- SHELL :
THERMOCOUPLES (4):
Tim
I.1.
TOITOTO
HOLDER -
-DESPOSITION PLATES
2 HALVES :
-HOLDER PIN
-PyC SHEETS
DEPOSITION END PLATES.
(4 LAYERS)
TOP AND BOTTOM -
PyC SHEETS,
TOP AND BOTTOM
Gas-Gap Fission Solid Capsule.!
is
1. Fig
.
RETURENCES
"
ssion-product release from
* Nuclear Safety,
:: 41 (1962), 35.
(2) CARROLL, R. M., KEAGAN, P. E., "Techniques for in-pile measurements of
fission-gas release," Nucl. Sci. Engag 21 ? (1965) 141.
[3] CARROLL, R. M., "Argon activation measures Irradiation flux continuously,"
Nucleonics 21 2 (1962) 42.
(4) CARROLL, R. M. and SISMAN, O., "fission-gas release during fissioning
of UO2," Nucl. Applications I 2 (1966) 142.
(5) REAGAN, P. E., MORGAN, J. G., and STSMAN, O., "F1881on-gas release from
pyrolytic carbon coated fuel particles during Irradiation at 2000 to
.: 2500T," Nucl. Sci. Engng. 23 (1965) 215-223.
[6] REAGAN, P. E., CARLSEN, F. I., and CARROLL, R. M. "Fission-gas release
from pyrolytic carbon coated fuel particles during irradiation, " Nucl. :.:
Sci: Engag. 16 3 (1964) 301-318.
(7) CARROLL, R. M., PEREZ, R., B., and SISMAN, O., "Release of fission gas
during fissioning of UO2," J. Am. Ceram. Soc. 48 2 (1965) 55.
(8) ROGERS, M. D., ADAM, J., "Ejection of atoms from uranium by fission
fragments," J. Nucl. Material 6 2 (1962) 182.
[9] CARROLL, R. M. and SISMAN, O., "In-pile Pission-gas release from single
crystal 102," Nucl. Sci. Engngo 21 2 (1965) 147.
(10) PEREZ, R. B., "A dynamic method for 19-pile libeion-sas release studies,"
Nucl. Applications 1 2 (1966) 151.
(11) CARROLL, R. M., PEREZ, R. B., and BTSMAN, O., "Release of flosion gas
i during fissioning of 002," J. Am. Ceram. Soc. 48 2 (1965) 55.
RWARENCES (Cont'a)
(12) PEREZ, R. B., NELSON, Paul, WATSON, G. M., privato commm.ication, oak
Ridge National Laboratory, March, 1967.
(13) REAGAN; P. E. et al., Reactor Chemistry Division Annual Progress Report
for Period Prading December 31, 1966, UBABC Report ORNL-4076, (1967) 99,
:. Oak Ridge National Laboratory, .. . : : ;
IT
.
..
7
.-
.
.
:
:
LIST A FIGURE CAPTIONS
Mg. 1. nov Dlageren of Mosion-Cas Rodense beperiment.
Mg. 2. Cross Section of Musica-Gas Release Capsule.
Mg. 3. Coated Particle Capoule for ORR C-1 hacility: Type 2.
Mg. 4. . Release Rate of
Kr from Mne-Grain Specimen (C1-12) of vog.
Mg. 5. Comparison of Kr Recoil Relouse Between single-Crystal and
Mao-Grain Vog' specimens of the same 81xe and Density.
: Mg. 6. Temperature Dependent Release Rate of
Kr from Mine-Grain UO . .
Detect-trar Characteristics.
Pg. 8. Comparison of Theoretical Model with Experimental Data.
.
Mg. 9. Release Rate Transfe:: Function vs irquenc
Mg. 10. Pyrolytic Carbon Coated Uranium Oxide Particles from Batch
. QR-348. Hag. 200%. (a) Vairradiated, (b) Irradiated to 9.4% burnup at
1350°C in capsule B9-27.
Mg. 11. Pyrolytic Carbon Coated Thorium-Uranium Carbide Particles
from Batch GA-310. Mag. 200x. (a) Unirradiated, (b) Irradiated to 10%
burnup at 1125C in capsule Cl-21.
::
. Mg. 12. Activity Bursts Indicate Failure of Individual Particles.
The bursts of activity were rich in gases with long half-lives, indicating
the gas bad been confined for a long time before release.
".
.
Mg. 13. Relationship Between R/B and Half Life for Fosion-Cas Re-
·louse from Particles with Draken Coatings.
.Me. 14. Relationship Between R/B and Half Life for Release of Inort
Cases iron Pyrolytic Carbon Coated Uranium Carbide Particles. ...
Me. 15. Cas-Cap Mosion salia Capoule.
END
DATE FILMED
5 / 18/67








VEN
K