JOTJRNAL
OF NUCLEAR
42 (1972)
MATERIALS
CREEP BEHAVIOR
OF CERAMIC
285-296.
0 NORTH-HOLLAND
NUCLEAR
fiir Material-
und
A
theoretical rate
between rate
estimation of
UO3
about
of
of
the
resulted and
due to the “thermal tracks.
1200 “C)
depend
and
and nitride To
of
Therefore,
on temperature
must
be markedly
this
in-pile
estimate creep
(below
lower
by
for
fuel
samples
In-pile
creep experiments
fuel temperatures
were
proportional
to
stress
(0 to
approximation,
4 kgf/mm3)
850 “C. Les it l’effort
p&s, la vitesse de la temperature
la porosite
aussi au taux de fission et se situent
(de 250’
essais
de
fluage
en
the in-pile (250-850
creep
of
rate is independent
inaccuracy,
d’UN
les vitesses infhrieures
de
de fluage
en accord
d’un
21celles observbes
pour
L’influence
de fluage
en pile
avec
une relation
creep
tests
on some
UN
samples
resulted Nach
The
strahlungsinduzierten
rate
liegt
for
porosity
UO3
under
dependence
found
relationship
to
comparable of
agree
the
with
for ceramic
conditions.
UN an
in-pile
creep
adequate
out-of-pile
einer bei
materials.
einer
UO3 induit de vitesses
thdorique
de la vitesse
par I’irradiation de fluage
sit&
conduit entre
de fluage
de
Uran-Spdtungsrate
nur
auf
die
“thermal
rods”
schwach
fragments
essentiellement
le
long
des
de fission. En con&quence,
fluage dependraient 1000 et
seulement
1200 “C)
Pour
conflrmer types
*
Presented
cette
de
mente
btre beaucoup
par
de fluage Science
November
oder
(unterhalb
liegen.
abgeschiitzten
zu iiberpriifen, der
sein
und Nitridbrennstoff
Ergebnisse wurden
entwickelt,
Durchmesser
von
durch
mehrere
I’exp&ience,
wurden
Divisions
2, 197 1. 285
bei
Joint
Fall
Meeting,
von
mit denen die Liinge
pneumatisch
Brennstofftemperaturen
und 850 “C durchgeftirt.
Experi-
Typen
belasteten
Brennstofftabletten kontinuierlich gemessen kann. Kriechexperimente unter Bestrahlung
en pile ont 8th and Nuclear
niedriger
die
in-pile-Kriechkapseln
St base de carbure
estimation
at the Basic
California,
Urn
les vitesses
et devraient
de capsules
des
peu de la temperature
plus faibles pour les combustibles et de nitrure. plusieurs
dG it des
trajectoires
Spaltsollten
Kriechgeschwindigkeiten
bedeutend
est
der
Demnach
temperaturabhiingig
de
f/mm 3. Ceci thermiques”
im Bereich
entlang
zuriickzufiihren.
1000-1200 “C) und fiir Karbid-
2 kg
1 x 1014
von 2 kp/mm3
Kriechgeschwindigkeit
pour un taux de fission de l.l014f/cm3. set et un effort “canaux
be-
in UO3
6 x 10-3/h und 8 x 10-3/h. Sie ist im wesent-
die bestrahlungsinduzierten
6.10-3/h et 8.10-3/h
der
von
und einer Spannung
fragmentbahnen
St un intervalle
AbschLtzung
Kriechgeschwindigkeit
die zu erwartende lichen
estimation
theoretischen
Spaltungen/cm3.s zwischen
Anaheim,
addquate
cbramiques.
of temperature
than
(entre
de
de UN
“C).
In-pile
Une
quelques sont
comparables.
sur la vitesse
in creep rates that are lower by an order of magnitude
was
limites
it 850 “C). pile
a BtB trouvhe
limits
dans le
Aux
pour les essais hors pile pour les mat&iaux
the
de fluage
0 et 4 kg f/mm3)
estimated
Within
sur
du combustible
rate (1 x 1013 to 2 x 1014 f/cm3. s) and are in the range before.
de
en pile
de fluage en pile eat indbpen-
UO3 dans des conditions
also to fission
mesurees
estimbes auparavant.
de grandeur
et
charges
vitesses
(entre
approximation
d’erreur
ordre
&re
de fluage
it des temperatures
250 et
de valeurs
Bchantillons
250 and 850 “C. The creep
peuvent
domaine
les
de longueur
combustible
1.10’3 it 2x 1014 f/cmS.sec)
Dans
at
de
Les essais
sont proportionnelles
dante
continuously.
on UO3 were performed
between
and, in a preliminary
by
change of pneumatically
can be measured
entre
et en premiere
several
developed
pneumatique
UO3 ont 6th realis&
carbide
les variations
d’&hantillons
continue.
sit&es
lOOO_
experiments,
capsules
loaded
are
along the
(de
which the length or diameter
rates
rods”
voie
manibre
creep rates should
W. Germany
Karlsruhe,
dans lesquelles
diam&re
par
fuel.
prove
types
de
*
197 1
realisees
range
a stress of 2 kgf/mm3.
It is essentially weakly
rate
6 x 10-3/h and 8 x 10-3/h for a fission
1 x 1014 f/cm3.s
fission fragment only
15 November
irradiation-induced
in a creep
AMSTERDAM
DIENST
Kernforrrchungszentrum,
Received
creep
and W.
Festkiirperforschung,
CO.,
FUELS UNDER NEUTRON IRRADIATION
D. BRUCKLACHER Inditut
PUBLISHING
werden an UO3
zwischen
250
Die Kriechgeschwindigkeiten American
Ceramic
Society,
286 sind
D. BRUCKLACHER proportional
4 kp/mmz) zur
zur
allgelegten
Uran-Spaltungsrate
tungen/cm3.s).
(1 x 1013 bis
Bereich.
Innerhalb
Kriechgeschwindigkeit
bestrahlung
(0
Niiherung,
bis
such
2 x 10’4 Spal-
Sie liegen in dem zuvor abgeschgtzten
ist
die
Spannung
und, in einer vorl&ufigen
der
nnabhgngig
Grenzen
der
Ungenauigkeit
unter
Neutronen-
von der Temperatur
(250 bis
AND W. DIENST In-pile-Kriechgeschwindigkeitsuntersuchungen einigen
UN-Proben
brachten
keiten, die urn eine dezimale sind als fiir UOZ unter keit
unter
Grdssenordnung
vergleichbaren
Fiir die Porosit,&tsabhiingigkeit
niedriger
Bedingungen.
der Kriechgeschwindig-
Neutronenbestrahlung
ubereinstimmung
an
Kriechgeschwindig-
wurde
mit einer entsprechenden
eine
gute
Beziehung
850 “C).
fiir keramische
1.
sideration 2). The thermal diffusion coefficient D was simply replaced by an irradiation-induced effective self-diffusion coefficient D * which was estimated from lattice point defect production or local overheating by fission spikes 1). This procedure resulted in useful creep rate values, but was not physically justified, because D is originally the product of the vacancy diffusion coefficient D, and the thermal vacancy concentration cr, th which generally cannot be replaced by the irradiation-increased vacancy concentration. But a slight adaption of the original model 3) rendered the above relation interpretable 4). Instead of vacancy production at grain boundaries in the original case one must envisage the annealing of irradiation point defects at defect clusters within the lattice. Annealing is favored at those clusters that are oriented to relax the operating stress by trapping vacancies or interstitials. Considering an interstitial cluster plane oriented perpendicular to a tensile stress direction one can see that the migration energy for transporting interstitials to the cluster edge is lower by u.Q than in the case of parallel orientation. The same is valid with an opposite sign for vacancies. By this energy change, a difference is caused in the flux of point defects which are annealed at clusters of different orientation. The flux difference can be formally described by a point defect concentration difference between neighboring clusters of different orientation which is approximately given by
Introduction
The specification of nuclear reactor fuel rods for high burnup must take into account mechanical interaction between fuel and cladding by the fission product swelling of the fuel and therefore requires sufficiently accurate knowledge about the swelling and creep behavior of the fuel. This was the reason for starting a Karlsruhe program for irradiation experiments on the in-pile creep of ceramic fuels. First a theoretical estimation was made of the irradiation-induced creep rates to be expected following considerations based on useful literature data of microstructure and density changes of clad porous UOZ after neutron irradiation l). The results indicated that below 1200 “C the creep rate of UO2 was increased markedly by irradiation, and should be expected to be between about 5 x IO-G/h and 2 x 10-4/h at a fission rate of 2.5 x 1014 f/ cm3 - s and a stress of 2 kgf/mmz *. It appeared to be only slightly temperature dependent. 2.
Theoretical estimation of irradiationenhanced creep rate
2.1.
ADAPTION OF THE MODEL FOR DIFFUSIONALCREEP
For the estimation mentioned rigorous use was made of the well-known relation for stressinduced diffusional creep
(1) (i’ creep rate, (T external stress, D volume selfdiffusion coefficient, 9 atomic volume, G grain size) which is valid for out-of-pile creep of UOz at the stress and temperatures under con*
1 kgf/mm2=1420
psi.
Mat,erialien ohne Bestrahlung
AC = cd2lkT
gefunden.
(2)
because oQ/kT << 1. In this relation c is no longer the thermal but the irradiation-induced point defect concentration at equilibrium be-
CREEP
BEHAVIOR
tween production Therefore,
OF
CERAMIC
NUCLEAR
FUELS
UNDER
NEUTRON
IRRADIATION
287
and annealing without stress.
eq. (1) which was derived
from the
same relation as given by eq. (2), but refers to thermal vacancy concentration at grain boundaries, can also be applied in the case of diffusion of irradiation-induced point defects, if G is replaced by an effective distance between of different orientation. Electron clusters microscopic observations of several authors on defect clusters and dislocation loops in ceramic fuels after irradiation point to a cluster density of about lOrs/cms in the temperature region concerned (T< 1300 “C). Therefore G was replaced by a “cluster cell” size of 1000 A. Finally one must take into consideration the fact that the density of sinks for point defect,s is much lower on the “cluster cell walls” than in the grain boundaries. Representing the cluster “cross section” by a strip 10 A wide along a loop 1000 A long, the degree of covering the cluster cell surface by point defect sinks was estimated to be l/100. Allowing for this factor and for G= 1000 A, eq. (1) is changed into L?no iirrad,= (I.5 X log CDle2) m D*. (3) In this, D * is the effective coefficient under irradiation. 2.2.
self-diffusion
IRRADIATION-ENHANCEDsm~-Dmmx3ro~ AND CREEP RATE
The way chosen for estimating D * has been described previously 134). Fig. 1 shows the D’ cd/s
I‘W llODlOaD 800 700 6&l
500
‘00r ’
Fig. 1. Effective uranium self-diffusion coefficient in UOs (partly also UN, UC) under irradiation at a fission rate of 2.5 X lOI f/cm3.s.
Fig. 2. Creep rate in UOz (partly also UN, UC) under irradiation at a fission rate of 2.5 x 1014f/cms.s and a stress of 2 kgf/mmz. results for a fission rate of 2.5 x 101% flcrn3.s and (formulating the annealing of point defects by cluster trapping) an effective dislocation density of lO*O/cmz. Different shaded areas represent the contribution of point defects from lattice collisions (“Frenkel-pairs”) annealing predominantly at clusters or by recombination, of thermal rods by ionization processes along the fission fragment tracks, and cf thermal spikes by spontaneous breakdown of displacement spikes. Fig. 2 gives the translation of fig. 1 by eq. (3) into irradiation-enhanced creep rates at a st’ress of 2 kgf/mma. ,4t higher temperatures the influence of irradiation is limited by the thermal creep rate. In UOz, an essential contribution from Frenkelpairs can be expected only between about 950 and 1300 “C. At lower temperatures Frenkelpairs are predominantly annealed by recombination which excludes a stress orientation effect. The decisive irradiation influence on the creep rate then turns to the thermal rods, this means channels of strong local overheating by the ionization processes along the fission fragment tracks. This conclusion is valid for oxide fuel, but cannot be transferred to ceramic nitride and carbide fuel. By their electronic conductivity thermal rods in these fuels will not become as hot because of (1) wide conduction electron-electron energy dispersion from the ionized channel along a fission fragment track,
288
D.
BRUCKLACHER
AND
W.
DIENST
(2) better heat conduction from a thermal rod to the surrounding lattice. Therefore,
the thermal
rod effect
on creep
rate must be strongly reduced compared to UOs, the reduction being limited by the range termed because an
“thermal spikes” in fig. 2. This is the thermal spike effect results from
immediate
transfer
of
the
displacement
(spike) energy to the lattice atoms and therefore should not show a large difference between oxide fuel and nitride or carbide fuel. It seems very difficult, especially with regard to the above point (1) (electron energy dispersion), to estimate the nitride and carbide creep rate level between the boundaries “thermal rods” and “thermal spikes”. But according to fig. 2 one must expect a steep creep rate increase with temperature above about 900 “C because the influence of Frenkel-pairs annealing at clusters is effective also in nitride and carbide fuel. To prove the thermal rod effect, which should only weakly depend on temperature, in-pile creep experiments at low fuel temperatures (about 400-700 “C) appeared very useful for the first series of experiments. However, irradiation temperatures must be increased up to about 1200 “C to study the competing influence of different mechanisms on the creep rat,e . 3.
3.1.
Irradiation capsules for in-pile creep measurements FR 2 CAPSULE
For the main series of in-pile creep experiments in the FR 2 reactor (Karlsruhe) an irradiation capsule was developed that can be used at fuel sample temperatures up to about 900 “C4~5). An axial section is shown in fig. 3. The sample consists of an alternating stack of UOs- and molybdenum (TZM)-rings, insuring a small temperature difference and a low crack sensibility (fig. 41. It is loaded pneumatically (O-4 kgf/mms) by a fixed helium pressure in a rather large pressure capsule which encloses two sodium-filled sample capsules with their
Fig.
3.
Axial
for continuous
section of a FR 2 irradiation capsule measurement
of the
sample
length
change by creep in ceramic fuels.
measuring systems. The sample temperature can be adjusted by varying the gap width between the sample and the pressure capsules. Temperature is measured by thermocouples in the sodium gap of the sample capsule and in the centering molybdenum tube of the sample stack. The sample length change is continuously measured by inductive
transducers
(differential
transformers) designed for a measuring range up to & 0.8 mm, maximum service temperature of 250 “C, and maximum fast neutron fluence of 3 x 1020 n/cm”. They are individually adjusted out-of-pile with regard to the influence of their temperature in-pile. Defects and deviations during irradiation can be detected and corrected by controlling and adjusting the sum of their secondary-coil voltages. After irradiation the results are corrected for the thermal expansion of the sample and the sample/measuring system holder. 3.2.
BR 2 CAPSULE
For in-pile creep experiments
at higher fission
CREEP
BEHAVIOR
Fig.
Creep sample stack of annular UOZ pellets (5 mm in diameter,
4.
OF
CERAMIC
NUCLEAR
FUELS
5.
X-ray
picture
of
the
irradiation
sample
length
capsule change
NEUTRON
IRRADIATION
1 mm thick) and molybdenum
289
rings.
picture is shown in fig. 5. This capsule should work at sample surface temperatures between about 600 and 1000 “C. The sample temperature can be varied during irradiation by changing the composition of a helium-neon mixture in the gas gap of the double-walled capsule. The
rates (up to about 4 x lOI f/cma.s compared to 1 x 1014fjcm3. s at the maximum in the FR 2 reactor) and burnups the capsule type CONFLUENT was designed and has been tested in a preliminary manner in the high flux reactor BR 2 (Mel, Belgium). An X-ray projection
Fig.
UNDER
“CONFLUEN’J!” by
creep
for continuous
in ceramic
fuels.
measurement
of the
D.
290
inner
capsule
containing
the
BRUCKLACHER
sample
AND
is filled
W.
DIENST
therefore
especially
useful
for
high-burnup
experiments. Moreover, there is much less adjustment required because of the absolute
with sodium-potassium. The sample has the same configuration and is loaded by the same principle as in the FR 2
length
capsule described above. However, the pressure on the sample can be changed during irradiation
pre-irradiation measurements of the friction loss in the loading system. The post-irradia,tion
in the
corrections
range
from
0.5
to
5 kgf/mms.
The
capability to run temperature and load cycles is very important because of the highly bimeconsuming type of experiment and the considerable microstructural variations in ceramic sample batches. The sample length change during irradiation is measured by means of electromagnetic microwaves. A resonant cavity with a movable piston-shaped bottom is fed by a klystron tube through a wave guide from the outside of the reactor. The piston is coupled with the sample. It is displaced with the sample length and simultaneously changes the volume and thereby the resonance frequency of the in-pile cavity. This is followed by comparing and adjusting equal out-of-pile equipment using a micrometer screw. Its position is continuously registered to plot the sample length change. This measuring system should be less sensitive to troubles caused by high temperatures and neutron fluences than inductive transducers and
Fig.
6.
Schematic
drawing
of
the
irradiation
3.3.
measurement.
SILOti
are similar
Some
care is needed
to those
on
in 3.1.
CAPSULE
A third capsule type VADIA was designed for continuous measurement of the diameter changes of short fuel pins under external gas pressure and neutron irradiation. Fig. 6 shows a schematic drawing. It is intended to investigate the creep behavior of cylindrical fuel samples under variable radial restraint at, representative temperature gradients. It can be especially useful to measure directly a “swelling pressure” of the fuel. Specimen length and diameter are about 100 mm and 7 mm, respectively. The cladding temperature is set between 600 and 700 “C and controlled by a gas gap sirnil% to 3.2. The sample capsule is designed to be pressurized up to about 150 atm. The pressure (external on the sample) can be changed during irradiation. The sample diameter change is continuously measured by the resonant cavity principle
capsule
“VADIA”
for
continuous
diameter change of fuel pin specimens under external gas pressure.
measurement
of
the
CREEP
BEHAVIOR
already
mentioned
OF
CERAMIC
in
3.2.
The
NUCLEAR
fuel
FUELS
pin
UNDER
NEUTRON
IRRADIATION
291
is
supported between two contacts at its mid plane, one of which is immovably connected with the bulk of the resonant cavity, being
placed
on a mobile
lever
the other
beam.
This
amplifies and transfers the sample diameter change, which is then turned into the axial direction
by
a lever
system
and led to the
piston in the resonant cavity. The lever system is kept under tension by the bellows between the resonant cavity and the specimen chamber.’ The pin is free to rotate about a horizontal axis to prevent blocking the contacts.
031
Fig.
8.
function
Creep rate of UOZ under irradiation of fission rate, standardized
as a
to a stress of
2 kgf/mm*.
4.
Experimental
4.1.
results and discussion
uoz
Compressive creep experiments according to 3.1 and 3.2 have been performed on UOa in the reactors FR 2 (Karlsruhe) and BR 2 (Mol) under the following conditions : fuel stoichiometry density : grain size :
: 2.000 < O/U < 2.005 96 * 1% TD lo-35 pm compressive stress : O-4 kgf/mm* 1.6 x 1013- 1.8 x 10’4f/cms.s fission rate : mean temperature : 250-850 “C uranium burnup : l-5 %. Compressive stress, fission rate, mean fuel temperature and uranium burnup were varied within the memioned Fig.
Fig.
7.
intervals.
7 shows the FR 2 results obtained
up
Strain rate of UOZ samples under irradiation
as a function of axial compressive stress, standardized to a fission rate of 1.2x 1014 f/cm3.s.
to now. The strain rates at a burnup of 0.2 to 0.3% each were plott#ed against the compressive stress. A standardization to a fission rate of 1.2 x 1Or4 f/cm3 . s was made, assuming the strain rate to be proportional to the fission rate. Positive strain rate values are due to the fission product swelling of the fuel. Apparently, there is a linear relationship between strain rate and stress which can be given by: daZ/dt= -(6x
lo-‘j/h.kgf.mm-*).(a-aa),
(4)
where ~0 is the swelling pressure of the UOa sample at which the linear swelling rate is just compensated by the fuel creep rate. According to fig. 7 the swelling pressure was ~0 = 0.8 + 0.15 kgf/mmz. The fuel swelling rate, which resulbs at o=O amounted to 0.8 vol. %I% burnup. The approximative proportionality between in-pile creep rate and fission rate which has already been applied above is confirmed by fig. 8. The measured points resulted from FR 2 and BR 2 experiments at fission rates from 1.6x 101s up to 1.8x 1014 f/cma+s. One point was added from an in-pile compressive creep test at a lower fission rate of 7 x lOi fjcm3.s which had been run at Harwelle). In fig. 9 the in-pile creep rat’es, standardized to a stress of 2 kgf/mm* and a fission rate of 1.2 x 1014f/cm3 . s, were plotted against the mean fuel temperature. The plot shows that the irradiation-enhanced creep rate is nearly independent of temperature.
D.
292 110-W
lO~Q3OOS~O
7pO
600
500
rgo
BRUCKLACHER
loo
200 *l
100
AND
W.
DIENST
initial strain rate decrease in samples of similar porosity is essentia,lly dependent on the product of fission rate and irradia.tion time. The decrease is possibly
due to a reduction
of fuel porosity.
Microstructure observations before and after irradiation have shown that the high initial strain rate is connected
with the disappearance
of very small fuel pores (diam.
< 1 ,um).
Fig. 9. Creep rate of UO3 under irradiation against temperature, standardized to a fission rate of 1.2 x 1014 f/cm3*s and a stress of 2 kgf/mm3.
Therefore, the in-pile creep rate of UOz at temperatures up to 850 “C is given by eirrad.= (0.35 + O.OB/kgf.mm-z)oR if
100
(5)
is the fission rate in f/U-atom. s. The burnup dependence of the creep rate under irradiation could be followed for a long time by a sample irradiated up to about 5% uranium burnup. Fig. 10 shows the sample length change as a function of the irradiation time. In the beginning the strain rate is high and decreases approximately by a logarithmic function up to at least 0.075-0.1% burnup. After a burnup of 0.2-0.3% the sample length change has apparently turned into stationary creep. A comparison with the Harwell creep
R
test, 6) at a lower fission rate suggests that the
Fig, 11.
200
3bo
COO 560
600 lrroddon
7bo I!me/Cycle
800 Ihl
Fig. 10. Strain of a UO3 sample as a function of irradiation time (0=4 kgfimm3, R= 1.2 down to 5 x 1013 f/cm3.s, T - 650 “C).
To control the continuous in-pile measurement of the sample length change, some postirradiation measurements were done using micrometers. All UO2 pellets and molybdenum rings of the creep samples had been individually numbered and measured before irradiation. After irradiation to about 1.5% burnup the length measurement was repeated for those pellets and rings that could be unmounted without damage from two irradiation capsules,
Length change of single UO3 pellets and molybdenum rings measured after irradiation to 1.5% burnup (Tsurrace = 450 W, R=7.2 x 1013 f/cm3*s).
CREEP
BEHAVIOR
OF
CERAMIC
NUCLEAR
FUELS
UNDER
NEUTRON
IRRADIATION
293
one of which was pressurized during itiadiation, the other irradiated at zero pressure. The results of the measurements
pl’esented in fig. 11 show
that, within the limits of accuracy,
the length
of unrestrained UOS pellets increased by swelling during irradiation, the length of UOS pellets under
load
decreased
by
cr’eep, and that
of
molybdenum rings remained unchanged. From these results a creep rate of 1.5 x 10-5/h (at a stress of 4 kgf/mmz and a fission rate of 7 x 1013 f/cm3 es) can be calculated, which is in good agreement with the foregoing in-pile measurements. 4.2.
UN
Some semi-quantitative in-pile creep data on UN and UC have been given in a previous report 4). The irradiation-induced creep rate of these samples was found to be lower by an order of magnitude than that of UOz under comparable conditions. The first in-pile creep test on a UN sample in a FR 2 creep capsule according to 3.1 has essentially confirmed that result. The sample consists of annular UN pellets and molybdenum rings and has been irradiated under the following conditions : fuel density : average grain size: UOz content : compressive stress : fission rate : mean t,emperature :
89 & 1%
TD
6-10 pm
6 vol. y0 4 kgf/mm2 3.6 x 1013 f/cms.s 750 “C.
The recorder plot has not shown a measurable sample length change. Therefore, considering a UN swelling rate of 1.3 vol. x/y0 burnup, it can be concluded that irradiation-induced creep is taking place at a creep rate of 1.5-2 x 10-6/h. To obtain a first rough knowledge about the influence of fuel porosity on the in-pile creep rate, two short UN fuel pins were irradiated, each of which contained several UN pellets of different porosity (83%87% TD). The fuel rods were 6.2 mm in diameter and were clad by 0.4 mm Inconel 625 or Incoloy 800 tube. The cladding has been isostatically hot-pressed on
Fig.
12.
UN
fuel pins,
different porosity
each containing
(83%87%
TD)
pellets
of
after irradiation
to
4 o/O uranium burnup.
to the fuel under high helium pressure. As a result, an immediate mechanical interaction was guaranteed between the cladding tube and the swelling fuel under irradiation. In-pile creep data were estimated from the cladding creep strength, the fuel swelling rate and the fuel diameter change, which resulted from very accurate measurements of the pin diameter (by profilometer) and of the gap width between fuel and cladding (by micrographs). The samples were irradiated to 4% burnup in a NaK-filled capsule at a mean fission rate of about 3.5 x 1013 f/cm3.s, cladding temperatures of 570 f 50 “C, and a maximum fuel temperature below 900 “C. Fig. 12 shows the two fuel pins after irradiation, the fuel density in the pins increasing in the upwards direction. The Inconel 625 cladding has been damaged by a longitudinal crack, apparently originating from the region of highest fuel density. The Incoloy 800-clad sample remained undamaged and could be used for postirradiation dimensional measurements. In fig. 13 the diameter change of this sample after irradiation was plotted over the fuel length with the fuel pellet density steps for comparison. The cladding strain decreases
D.
294 ADiD
BRUCKLACHER
AND
W.
DIENST
I%1 20
15 I
Fig.
13.
Fuel
diameter
increase
over
(83~87%
fuel
length
of
a UN
TD) after irradiation
the fuel density. Assuming the fuel swelling rate to be independent of porosity (this is justified under the irradiation conditions prevailing), one can conclude that there is a clear porosity dependence of the in-pile fuel creep rate under cladding restraint. The pressure load on the fuel by the cladding restraint was estimated to be 1.5 kgf/mms. From fig. 13 and a fuel swelling rate of 1.3 %I% burnup, UN in-pile creep rates of about 1.8 x IO-s/h, 1.32 x lo-a/h, 1.17 x 10-s/h, and 9.2.10-7/h result for the different densities of 83, 84.1, 85.3, and 86.4% TD, respect,ively. with
The proportions between these creep rates fit the empirical curve in fig. 14, which was
pin
containing
pellets
of
different
porosity
to 4% uranium burnup.
obtained from out-of-pile creep data of other authors on UOz, (U, Pu)Oz and A1203 samples of different porosity (1-1Oo/o). This curve can be approximated by the equation ipI&= 1 + 0.125 P2,with the porosity P given in vol. %. Additionally, the diagram shows the in-pile creep rate proportion of 83.50/O and 94% dense (U, Pu)N which was estimated from the abovementioned nitride fuel swelling rate and the fuel volume increase measured after BMI fuel pin irradiations 7). The temperature dependence of the cladding restraint was taken into consideraGon. If the curve in fig. 14 is applied to the UN in-pile creep data, the creep rate range marked
.
loI /
lh,swork
0 BHI
3
1 i/
I 0
Fig.
14.
Creep rate in ceramic Al202
out-of-pile
materials data,
30
20
IO
as a function
UN
in-pile
of porosity
creep
rate
P
obtained
proportions
fitting
from
UO2,
the curve.
(U,
Pu)Oz
and
CREEP
BEHAVIOR
OF
CERAMIC
ROCLEAR
Fig. 15. Range of me~ured in-pile creep rates of UOs and UN samples, standardized to a fuel density of 96% TD, a fission rate of 2.5~ 10’4 f/cm*-s and a stress of 2 kgfjmma, in comparison with the result of foregoing theoretical sstimations.
fig. 15 results for 96% dense UN, which can be compared with the range resulting for uoa. in
5.
Conclu~ons
The following conclusions can be drawn from the reported results : Ceramic nuclear fuels show measurable fissioninduced creep rates down to fuel temperatures of about 250 “C. It is possible to measure these creep rates cont,inuously in-pile. The measured in-pile creep rates are in agreement with foregoing theoret.ical estimations (fig, 15). Especially, the ~ssioIl-induced creep rate proved to be approximately inde~ndent of the oxide fuel temperature between about 250 and 850 “C. This result confirms the dominant influence of the thermal rods along the fission fragment tracks which was suggested by theoretica. considerations. The same is valid with regard to the in-pile 0ree.prates of UN samples : These creep rates were found to be about an order of magnitude lower compared to oxide fuel, probably because of a strongly reduced thermal rod effect. The in-pile creep rate of UOa is proportional to the operating stress.
FUELS
UNDER
BEUTRON
IRRADIATION
295
- The in-pile creep rate of UOa appears to be approximately proportional to the fission rate. - The in-piIe creep rate of high-density UOa considerably decreases up to a buwup of about 0.1%. Therefore, a comparison of creep rates from various experiments should only be made after burnups of at least 0.3%. The initial decrease of t#hein-pile creep rate is possibly due to fuel perosity changes. - The porosity dependence of the in-pile creep rate can be probably taken from adequate out-of-pile relationships. The results are consistent with those of other authors 8-n). BMI experiments 10) were performed at higher irradiation temperatures and have shown that in the temperature range of about X000-1200 “C the creep rate of UOa is still enhanced markedly by irradiation, but is highly dependent on the irradiation temperature. Future experiments in Ka~lsruhe are planned to study the in-pile creep behaviour of (U, PufOa, the influence of fuel porosity and of high burnup. However, no cause has been found up to now to consider and search for an essential influence of the fuel grain size. Acknowledgements The authors thank H. Hlfner, W. Neumann, and H. Will for design and construction of creep capsules. Thanks are also due to X. ~lumhofer, W, Hartlieb, and K. Philipp for assistance in operation of creep capsules, and to H. Enderlein, R. Pejsa, and F. Weiser for the post,-i~adiation examinations. References D. Brucklacher, W. Dienst and F. Thiimmler, Considerations on the Creep Behavior of UOs under Neutron Irradiation, Report KFK-817, 1968 and Ceramic Nuclear Fuels, Proc. of an Intern. Symp., Nucl. Div. Amer. Ceram. Sot., 1969, Special Publ. No. 2, p. 169 A. R. Wolfe and S. F. Kaufman, Mechanical Properties of Oxide Fuels, Report WAPDTM 587, 1967 F. R. N. Nabarro, in: Report on a Conf. on Strength of Solids (Phys. Sot., London, 1948) D. Brucklacher, W. Dienst and F. Thiimmler,
D.
296 Investigations
on Cleep of Ceramic Nuclear Fuels
AND
8)
W.
D.
DIENST
J.
Clough,
Irradiation
Induced
under Neutron Irradiation, in : Fast Reactor Fuel
Ceramic Fuels, in: Fast Reactor
and Fuel Elements,
Elements,
Proc.
Karlsruhe,
1970) p. 321
(GfK, 5)
BRUCKLACHER
Il.
Karlsruhe,
Brucklacher
36 (1970)
Proc. of an Intern. Meeting
1970) p. 343 and W.
7)
W.
M. Pardue,
communication,
A. A. Bauer and D. L. Keller,
of Mixed Nitride
(U, Pu)N
as a Fast
Symp.,
Nucl.
1969, Special Publ.
Div.
No.
Amer.
2, p. 116
Ceram.
Meeting
of
(GfK,
E. C. Sykes and P. T. Sawbridge, The Irradiation Dioxide,
CEBG
Report RD/
B/N 1489, 1970, also private communication,
1971
Fuel, in: Ceramic Nuclear Fuels, Proc.
of an Intern. Sot.,
9)
Creep of Uranium
D. J. Clough, private
Reactor
J. Nucl. Mater.
244
6)
Potential
Dienst,
of an Intern.
Creep
Fuel and Fuel
10)
J. S. Perrin, J. Nucl.
ii)
A. A. Solomon and J. L. Routbort, at
ACS
Meeting, 1970
Basic
Mater.
Science
Gatlinburg,
and
39 (1971)
Presentation
Nuclear
Tennessee,
1970
175
Divisions
November
5,