Journal of Nuclear Materials 65 (1977) l-8 8 North-Holland Publishing Company
IRRADLATION INDUCED CREEP OF CERAMIC NUCLEAR FUELS
W. DIENST Kemforschungszentrum Karlsruhe, Institut fUr Material-und Festktirperforschung, Federal I&public of Germany
This contriiution gives a review of the experimental results and accompanying theoretical considerations. The mechanisms considered for irradiation creep are: relaxation of elastic stresses by fission spikes, promotion of dislocation slide by thermal spikes, preferential, stress-orientated nucleation of dislocation loops and preferential growth of dislocation loop. A survey over the irradiation creep rates attributed to steady-state creep shows eirr 5 u* F for oxide fuel in the stress and fission rate ranges of o = lo-50 MN/m2 and F = 3 X 1012-L X 10 l4 f/cm3 - s at burnups <3%. There seems to be a continuous increase of the irradiation creep rate of oxide fuels with incmashrg temperature. However, that increase cannot be directly interpreted through a thermally activated process. It seems that the irradiation creep rate will also depend on fuel porosity, on plutonium distribution in mechanically blended UO2PuO2, but not substantially on the plutonium content per se. Some results were already given for carbide and nitride fuels, which show the irradiation creep rate to be lower by about a factor of 10 than for oxide fuel under comparable conditions. Primary irradiation creep has been observed up to (3-S) X lOI f/cm3 and could prevail up to 1 X 10” f/cm3. Cette contribution donne un rCsumC des r6sultats exp6rimentaux avec les considhrations thhorlques. Les micanismes consid&& pour le fluage sous irradiation sont: relaxation des contrahttes Clastiques par les pointes de fission, promotion du glissement des dislocations par pointes de feu thermiques, germination prhfhrentielle orientie par la contrainte de boucles de dislocations et croissance prkfirentielle de boucles de dislocations. Un examen des don&es sur les vitesses de fluage sous hmdiation dans le stade secondaire montre que I& * IJ. F pour le combustible oxyde darts les intervalles de contrainte et de vitesse de fission de u = lo-50 MN/m2 et F = 3 X 1012-1 X 1014 f/cm3. s pour des taux de combustion <3%. I1 semble qu’il y ait une augmentation continue de la vitesse de fluage sous irradiation des combustibles i oxyde avec la tempirature croissante. Cependant, cet accroissement ne peut pas Btre interpr6ti directement par un processus acti& thermiquement. 11semble que la vitesse de fluage sous irradiation d6pende aussi de la porosite du combustible, de la distribution du plutonium dans l’oxyde mixte UOz-PuO2 obtenu par m&nge m6canique, mais non de fa9on substantielle de la teneur en plutonium en elle-m5me. Quelques rksultats ont deja et6 don&s pour les combustibles carbure et nitrure, qui montrent que la vitesse de fluage sous irradiation est inferieure d’un facteur d’environ 10 $ celle d’un combustible a base d’oxyde sous des conditions comparables. Le fluage prlmaire sous irradiation a Qt6 observh jusqu’i (3-S) X 1019 f/cm3 et pourrait p&valoir jusqu’a 1 X 1020 f/cm3. Dieser Beitrag glbt einen Uberblick tiber die experimentellen Ergebnisse und begleitende theoretische Uberlegungen. Diese ijberlegungen betreffen die folgenden Mechanismen fur das Bestrahlungskriechen: Relaxation elastischer Spannungen durch fission spikes, Forderung des Versetzungsgleitens durch thermal spikes, spannungsorientierte Kelmbildung ftlr Versetzungsschleifen und spannungsorientiertes Wachstum von Versetzungsschleifen. Em Uberblick tiber die Bestrahlungskriechgeschwindigkeiten, die auf stationares Kriechen zurtickgeftihrt werden, zeigt, dass fur Oxidbrennstoff bei Abbriinden <3% die Kriechgeschwindigkeit Cirr proportional (I*F ist, wenn die Spannung im Bereich (I = lo-50 MN/m2 und die Spaltungsrate im Bereich F = 3. 10r2-1 - 1014 Spaltg./cm3 0s liegen. Die Bestrahlungskriechgeschwindigkeit von Oxidbrennstoff scheint mit steigender Temperatur kontinuierlich zuzunehmen. Diese Zunahme kann jedoch nicht direkt durch einen thermisch aktivierten Prozess gedeutet werden. Es scheint such eine Abhiinglgkeit von der Brennstoffporositiit und von der PuVerteilung in mechanisch gemischtem UO2-PuO2 zu bestehen, aber nicht vom PuGehalt an sich. Es gibt such schon einige Ergebnisse ftir Karbid- und Nitridbrennstoffe, nach denen die Bestrahlungskriechgeschwindigkeit etwa urn den Faktor 10 niedriger lie t als fur Oxidbrennstoff unter vergleichbaren Bedingungen. Prim&s Bestrahlungskriechen wurde bis zu 3 bis 50 10’ 8 Spaltg./cm3 beobachtet und kiinnte bis zu 1. 1020 Spaltg./cm3 vorherrschend sein.
1. Introduction
cladding contact. Substantial plastic strain of the cladding has been found in fuel rod irradiation tests, for oxide fuel as well as for carbide and nitride fuel. To analyse such results and to make a safe fuel design, in-
In principle, fuel swelling from fission product accumulation’ can cause cladding deformation on fuel/ 1
W.Diem /Irradiation induced creep of ceramic nuclear fireIs
2
pile measurement of the irradiation creep rate of the fuel materials is required. For there were several rather early indications of considerable irradiation-induced plasticity in the temperature range where thermal creep of ceramic fuel is negligible [l] . The contact pressure between fuel and cladding and the cladding strain depend on the ratio of swelling rate to creep rate of the fuel under the cladding restraint. Of course, that is only true, if there is sufficient void volume in the fuel pin to accomodate the fuel swelling. However this is the normal case, in particular due to the fuel porosity chosen in consideration of the burnup goal. On the other hand, the porosity dependence of irradiation creep will thereby play an important role. Considerations on possible mechanisms for irradiation-induced creep soon resulted in the conclusion that the electronic conductivity in ceramic fuels should keep the irradiation creep rate by thermal spikes relatively low. Therefore, a strong interest arose to compare the irradiation creep of oxide fuel and carbide or nitride fuel. 2. Physical mechanisms for irradiation creep
\ Neutron
Structural materiol fuel
u
u
Fig. 1. Relaxation of elastic stress in a fission spike. Arrows indicate stress, broken lines indicate initial specimen shape, solid lines indicate final specimen shape (from [ 131).
2.1. Relaxation of elastic stresses
The first proposals for an irradiation creep mechanism dealt with the relaxation of the elastic stresses in the crystal lattice volume temporarily disarranged by a fission spike [2,3] (fig. 1). The corresponding considerations were made with regard to metallic fuel, but can be readily applied also to ceramic fuel [4]. The irradiation-induced creep rate results to be 4rr -fi-
u/E,
where u is the stress, E the Young’s modulus, and fi the fission rate. This expression can give results in reasonable agreement with experimental measurements. However, the spike volume assumed can be varied rather widely and the degree of relaxation is uncertain.
t iii
2.2. Enhancement of dislocation slide by thennal spikes 0
Other considerations dealt with the purely thermal effect of fission spikes with regard to their influence on dislocation movement.
+m
21,
3+m
4tm
Stm
61,
TilllO
Fig. 2. Maintenanceof primary creep by fission spike annealing (from [S]).
3
W.Dienst/Irradiationinducedcreep of ceramicnuclearfuels
In one case it is a matter of annealing recovery of stressed fuel material in every thermal spike volume. By this process the fuel, being strained by thermal creep, is kept in the primary creep phase by irradiation influence [5] (fig. 2). The creep rate is &1rr.= eprim therm related to a certain stage of primary creep, which is determined by both temperature and fission rate. Only thermal creep can be enhanced by that mechanism, which, therefore, is restricted to accordingly high temperatures. It appears difficult to give a sufficiently accurate formulation of primary creep, due to the very short start-up period during which irradiation creep measurements can be compared to calculations by the above expression. In the other case thermal spikes were considered to generate strong thermoelastic shock-waves [6] . These shock-waves could interact with dislocations and, by this means, enhance dislocation movement. However, there is no discussion of a corresponding quantitative correlation.
2.3. Stress-otiented fomtation and growth of dislocation loops While the above-mentioned mechanisms are based on a random disarrangement of the crystal lattice within the fission spikes, other mechanisms considered depend on the formation and growth of dislocation loops under irradiation. To be effective for irradiation creep, these processes must show a preferential orientation relative to the direction of the applied stress. Then one has to distinguish 1) stress-oriented preferential formation of dislocation loops in point defect cascades [7,8] (fig. 3), and 2) preferential growth of appropriately oriented dislocation loops by interstitial capture [9] (fig. 4). Both processes were also considered simultaneously [lo] ; then the stress-oriented loop formation takes place by the condensation of point defects in “thermal rods” along the fission fragment tracks. In any case the irradiation-induced creep rate is
t --------------
1 I I
I I Loops Interstitial 0
0 , Groins Fig. 3. Stress-orientedpreferentialformation of dislocation loops in point defect cascades (from [ 131).
obove below
concentration I,
meon interstitial
I I
(bias) ot loops diffusion
to capture
II J
Fig. 4. Preferential growth of appropriately oriented dislocation loops by interstitial capture.
W.Dienst / Irradiation induced creep of ceramic nuclear fuels
4
The consideration on stress-oriented growth of dislocation loops are closely related to those about irradiation-induced creep of non-fissionable materials. It appears obvious to consider the energy balance for the capture of irradiation point defects at dislocation loops of different orientation relative to the stress direction. This procedure gives rather simple results and a quantitative correlation to irradiation-induced diffusion. However, in reality point defect annealing cannot result in point defect concentration gradients between dislocation loops of different orientation, because the capture probability does not depend on the energy gain to be realized. Therefore, one must rather consider the long-range interaction forces between point defects and dislocation loops, which result from their stress fields. The conclusions as for preferential growth of dislocation loops could be similar [ 11,121. Finally, a further mechanism for irradiation creep appears worth citing from a complete review [ 131 : diffusional creep by vacancy diffusion between dislocation loops (of different orientation relative to the stress direction), which are formed under irradiation
r
1
I I I I I
I
I
I I
L,
\, V,
I
LOOPS Vacancy
Horczontal vertical
loops loops
grow;
shrink
I I
.-I
Fig. 5. Diffusional creep by vacancy diffusion between dislocation loops formed under irradiation (from [ 131).
(fig. 5). It is a variety of Nabarro-Herring creep, which comes into question only for high temperatures and can be considered an alternative compared to the “continuous primary creep” treated in section 2.2. The review article mentioned [ 131 is a more corn.. prehensive presentation of irradiation creep than the short version given in the above sections. 2.4. Effectiveness of the mechanisms TO discuss the real effectiveness of the different mechanisms considered for irradiation creep of ceramiC fuels, some more general experimental observations shall be anticipated. Several times an incubation period of relatively low, increasing creep rate occurred prior to the true primary creep of decreasing rate (“sigmoidal” creep curve) [ 14,16,17]. That incubation period does not fit the relaxation of elastic stresses by fission spikes, which should operate from the irradiation start, but it suggests a dependence on dislocation structures formed under irradiation only. In this connection the incubation period contradicts the stress-orientated formation of dislocation loops in point defect cascades. The large difference between the irradiation creep rates of oxide and carbide or nitride fuel suggests a thermal spike effect. The dependence of the creep rate on stress and fission rate is not suited for discrimination, because in most cases &. u u . P . The temperature dependence is rather uncertain, however, it seems to fit the idea of a thermal spike effect. The most probable mechanism is a combination of dislocation loop growth, possibly also dislocation climb in a dislocation network, with a thermal spike effect. This is entirely true for relatively low temperatures [ 141. At higher temperatures dislocation loop growth could occur alone, in particular above the temperature threshold for considerable vacancy migration in the crystal lattice, which is at 600-7OO’C for ceramic fuel. At very high temperatures a combination of the irradiation-induced processes with those of thermal creep will occur. It seems particularly problematic to make statements on the temperature dependence of the irradiation creep rate. However, the experimental results will show, that the problem is more of theoretical than of technical significance.
W.Dienst 1 Irradiation induced creep of ceramic nuclear .fuels
3. Measuring methods The experimental methods used to measure irradiation creep of ceramic fuels are only briefly mentioned in the following. Of course, they will be treated in detail by special contributions. To some extent, there are strong similarities with the measuring devices for non-fissionable materials. In-pile creep capsules were designed for torsion tests on helical spring-shaped specimens [ 15 ,161, bend tests on small rectangular section bars [ 17,181 , compression tests on annular pellets or stacks of fuel rings [ 19-251 , and for a tensile test (on a UC sample) only in one case [ 171 . At particularly low temperature and/or high fission rate Na or NaK was used as a heat transport medium for improved heat transfer from the fuel sample. Considerable computation work was done to analyse the temperature distribution in the sample and its influence on the evaluation of measured strain rates [23,26,27] . The specimens were creep-tested under deadweight load [15], spring load [ 16,171 or gas pressure load [20,21,24,25] . T,he specimen length change was measured in-pile by a differential transformer [15,16,20, 24,251, pneumatic transducer [2 l] , microwave resonant cavity [20] , or remote micrometer [ 171, or outof-pile by neutron radiography and post-irradiation optical measurement [17,18]. The compression tests meet the problem of fuel/ cladding interaction in the best way, but the torsion tests allow a more detailed examination of the creep curve. A particularly efficient measuring method seems to be given by variable, controllable gas pressure load and microwave resonant cavity measurement for specimen length change.
4. Experimental results 4.1. Parameter dependence of steady-state irradiation creep 4.1.1. Dependence on stress and fission rate Concerning the oxide fuels UOZ and UOZ-PuOs , there is a general agreement, that the irradiation creep rate is proportional to the stress
5
in the range of u = lo-50 MN/m2, and at low burnup. In a single test on U02 at higher burnup (>3%) a stronger stress dependence was indicated 1281 about @in.
_
a1.5
.
For carbide and nitride fuel it seems possible, that the stress dependence of irradiation creep is also given by GK. - u. At any rate, if etr, - an is correct, then n appearstobe
For UO2 this relation proved valid in the range @= 3 X 1012 - 1 X 10’4f/cm3 - s [14,20]. For U02-Pu02 it appeared applicable within the limits of measuring accuracy, in the range of $ = 7 X lOI - 1.2 X 1014 f/cm3 * s. Therefore, there is no objection to proportionality. Only at high temperatures >- 11 OO’C do the creep rate results of different authors for U02 [ 18,2 1,241 differ so much, that the fission rate dependence appears to be uncertain, though the irradiation-induced increase of the creep rate was also assumed to be proportional to the fission rate [22,24]. The irradiation creep rates of UN [26,28], UC [ 181, and (U, Pu)C were measured only at constant fission rates (5 X 1013, 7 X lo’* and 1 X 10L4f/cm3 * s). Therefore it was not yet possible to verify the fission rate dependence. To study other parameter dependences, all experimental results for the irradiation creep rate of ceramic fuels were standardized using pirr. -0.b
and are presented in figs. 6 to 8, which show the creep rate versus the sample temperature. Fig. 6 gives the results for UO2, fig. 7 for UO2 -Pu02, and fig. 8 for UN, UC and (U, Pu)C in comparison with U02. The first measurements on nitride and carbide fuel samples resulted in irradiation creep rates that are lower by a factor of about 10 than for oxide fuel under standardized conditions. Such a trend had been expected according to theoretical considerations [ 10, 201: the efficiency of the “thermal rods” along the fission fragment tracks will be reduced by the electronic conductivity of the fuel. The reduction results from a wide energy dispersion from the ionized channel
W. Dienst /Irradiation
6
induced creep of ceramic nuclear fuels
10-6.
10-7.
10-7
0.6
.
. IO
14
.
. \. 1.6
2.2
Fig. 6. Correlation of irradiation-induced (pellet density 96-9870 TD).
10-E
1
1.6
3.0
I/l,
06
IO-3K-1
creep rates for UO2
along the fission fragment track and afterwards from a good heat conduction from the “thermal rod”. 4.1.2. Temperature dependence Figs. 6 and 7 show a distinct, continuous increase of the irradiation-induced creep rate of oxide fuel with increasing temperature. This differs from previous assumptions, that irradiation creep is athermal in a wide temperature range up to about 1000°C. The flattest and the steepest creep rate increase, that can be assumed, are plotted in fig. 6. The steep increase has the advantage to have been measured on a single U02 sample by gradual temperature change. Moreover, fig. 7 shows, that the U02-Pu02 results correspond better with the steeper temperature dependence. However, the increase appears to be curved in the Arrhenius
06
12
IO
Fig. 8. Irradiation-induced fuel compared to UO2.
14
1.6
l/T.
10-3K-’
creep rates of carbide and nitride
plot so that no definite thermally activated process can be attached. The temperature dependence observed could rather be explained by the (mean) sample temperature contributing to the temperature distribution in the thermal spikes, which are then considered to be the decisive mechanism for irradiation creep. Fig. 8 shows, that there is no indication of a similar temperature dependence for UN and UC up to now. This may be due to the minor role of thermal rods in these fuels, which has already been mentioned in the foregoing paragraph. In view of the considerable scattering of the measured values shown in fig. 6 it is recommendable to assume also the irradiation-induced creep rate of oxide fuel independent of temperature, for the purpose of fuel pin design and modeling calculations [20] : Eirr. = c1 YYP:,
_
1200
lo-3
.
ID00
600
700
.
600
5qo
.
400
3y
I. ‘C
A
3
10-4.
u is the stress, and i the fission rate. It makes no sense to only formally consider thermally activated irradiation creep processes. For irradiation-enhanced creep of oxide fuel at high temperatures the expression where
m-5.
&in.. = (q t c3 *I$*
10-6.
was suggested [22] . As mentioned
0.6
00
10
Fig. 7. Irradiation-induced for comparison.
creep rates of UO2-PuO2, and UO2
exp(-Q/RT)
,
above, the fission rate dependence seems to be uncertain in this case, from theoretical considerations of the creep mechanism as well as from the experimental results. Therefore it is rather recommended at present to assume only thermal creep at high temperatures.
W.Dienst / Irmdintion induced creep of ceramic nuclear fuels
4.1.3. Porosity dependence
Without doubt the dependence of the irradiation creep rate on the fuel porosity is of considerable technical importance with regard to fuel pin performance. However, only little work was done to examine the porosity dependence,and in particular for UOa only high-density samples of 96-98% TD were tested. Concerning UOz -PuOa there are results for more porous samples of 86 to 95% TD. Fig. 7 shows a distinct increase of the irradiation creep rate with increasing fuel porosity. However, for the values measured in Karlsruhe (Brucklacher, Dienst) that increase is additionally influenced by the rather inhomogeneous Pudistribution in the mechanically blended UOz-PuOz fuel, which is demonstrated from the small differences between the values measured on 93.5 and 86% TD samples. Nevertheless, the expression previously sug gested for the porosity dependence of the irradiation creep rate [20] P, = C,,(l + p2/8) , where P is the fuel porosity in vol.% turned out to be still applicable, under the assumption, that the PuO? particles act as pores (because of their high irradiationinduced plasticity) proportional to their total volume contribution [28,30] . Then the effective fission rate in the UOa-matrix of the fuel, which is mainly determined by fission fragments from the PuOz particles, resulted to be l/3 of the mean fission rate in the UOzPuOa fuel. The good correspondence of all values measured on UOa PuOa samples at high temperatures is probably due to the irradiation-induced densification of the porous fuel and homogenization of the Pu-distribution in the fuel [30] . The present measurements for UN (93% TD), UC (97% TD), and (U, Pu)C (84 and 96% TD) are not yet appropriate for an assessment of the porosity dependence of irradiation creep in such fuels. 4.1.4. Influence of the &-content At first the measuring values in fg 7 gave the impression, that irradiation creep of UOa -PuOz is faster than for UOa . However, under the assumption mentioned in the foregoing paragraph the differences in creep rate can be explained by the influence of fuel porosity and Pu-distribution. Therefore there is no reason at present to consider an influence of the Pucontent per se.
1
4.1.5. Influence of grain size There is no indication of a considerable influence of the fuel grain size on irradiation creep [29] . Irradiation creep measurements on UOZ samples of about 10 pm and of about 25 E.trngrain size [ 14,20,21] did not result in any systematic difference. This is not surprising, because the irradiation creep rate is probably determined by disturbed zones or by dislocation structures, the dimensions of which are below the grain size by one order of magnitude. 4.2. Primary irradiation creep Rather few results have been reported dealing with primary irradiation-induced creep. On the other hand certain measuring values could be decisively influenced by primary creep, without being stated to be so [ 16, 241. The data available [14,18,28] show that primary creep appears prevailing up to fission densities of (3-S) X 1019f/cm3, maybe even up to 1 X 10zof/ cm3. The time dependence of primary creep strain for oxide fuel was found to be m Eilr.prim,-t , m = OS-O.65 , eirrprhn. coming up to the order of I-10%.
The time dependence given above is similar to that of out-of-pile primary creep. The exponent m is independent of the stress u and also independent of the fission rate P in-pile [3 l] . If the load on the sample is increased, after steady-state creep has been reached, primary creep will be observed again. It can be concluded from these observations, that even under irradiation time dependence of primary creep of oxide fuel is mainly determined by the creep deformation, the irradiation influence being of minor importance. Since the total strain of the fuel in fuel pins will amount to some percent only, it could be entirely due to primary creep [ 14,3 1,321 . In the expression errrprim. = Kt" , K will yet depend on the stress, the fission rate, the fuel porosity, and the temperature, K = f(o, k, P, T). The stress dependence seems to be K - u as in the case of steadystate creep [32]. 5. Importance of further experiments
Due to the irradiation creep rates measured on oxide fuel, the cladding stress caused by steady-state
8
W.Dienst / Irradirrtioninduced creep of ceramic nuclear fuels
swelling of the fuel in oxide fuel pins can be expected to be too low to play a crucial role, as long as suffkient internal void volume is available. Other processes than steady-state fuel swelling have to be considered to cause the important plastic deformation of fuel pin cladding tubes, which did occur by mechanical interaction between fuel and cladding. Those processes must imply short-time straining, but could be still influenced by primary irradiation creep because of the high primary creep rate. Therefore further experiments on the primary irradiation creep of oxide fuel seem to be of considerable interest. On the other hand, there is no technical need for continuing to test the parameter dependence of the steady-state irradiation creep of oxide fuel. However, for carbide fuel pins (and perhaps nitride fuel pins) the cladding strain by steady-state fuel swelling will decisively depend on both primary and steadystate creep of the fuel. In this case it has to be analysed, whether the fuel swelling can be largely accomodated in the internal void volume under the restraint of a high-strength cladding. Therefore the examination of carbide fuel (particularly (U, Pu)C) irradiation creep is still important. On this occasion primary creep is to be properly recorded and analysed. The parameter studies should focus attention on the porosity dependence of carbide fuel irradiation creep.
References [l] D. Brucklacher, W. Dienst and F. Thlimmler, Rep. KFK 817 (1968) and Ceramic Nuclear Fuels, Proc. Int. Symp. Nucl. Div. Amer. Ceram. Sot., Spec. Publ. No. 2 (1969) 100 (21 ST. Konobeevsky, Soviet Journal of Atomic Energy 9 (1961) 707 [3] J.A. Brinkmann and H. Wiedersich, ASTM Spec. Tech. Publ. No. 380 (1965) 3
(41 R.G. Anderson and R. Hargreaves, TRG Rep. 1975 S (1973) [5] W.R.D. Wilson, J.A. Walowit and J.N. Anno, Rep. BMI 1857 (1969) C20 [6] H. Blank, Phys. Stat. Sol. 10 (1972) 465 [7] R.V. Hesketh, Phil Mag.7 (1962) 1417 [8] R.V. Hesketh, Rep. BNL 50083 (1967) 389 [9] M.F. Ashby, Scripta Metall. 6 (1972) 1231: and Harvard Univ. Rep..no.4 (1971). ’ 1101D. Bruckiacher, W. Dienst and F. Thiimmler, Fast reactor fuel and fuel elements, Proc. Intern. Meeting, GfK, Karlsruhe, 1970, p. 343 IllI P.T. Heald and M.V. Speight, Phil. Mag. 29 (1974) 1075 1121R. Bullough and J.R. Wiiiis, Phil. Mag. 31 (1975) 855 1131E.R. Gilbert, Reactor Technol. 14 (1971) 258 iI41 A.A. Solomon, J. Amer. Ceram. Sot. 56 (1973) 164 [ISI A.A. Solomon and R.H. Gebner, Nucl. Techn. 13 (1972) 177 1161E.C. Sykes and P.T. Sawbridge, Rep. CEBG-RD/B/N 1489 (1969) (171D.J. Clough, Fast reactor fuel and fuel elements, Proc. Intern. Meeting, GfK, Karlsruhe, 1970, p. 321 [I81 D.J. Clough, J. Nucl. Mater. 56 (1975) 279 [ 191 D. Bruckiacher and W. Dienst, J. Nucl. Mater. 36 (1970) 244 [20] D. Brucklacher and W. Dienst, J. Nucl. Mater. 42 (1972) 285 [21] J.S. Perrin, J. Nucl. Mater. 39 (1971) 175 [22] J.S. Perrin, J. Nucl. Mater. 42 (1972) 101 [23] J.L. Routbort, Rep. ANL-RDP9 (1972) 6.8 [24] I.S. Golovnin et al., Fuel and fuel elements for fast reactors, Proc. Symp. Brussels (IAEA, Vienna, 1974) Vol. I, p. 185 [ 251 F. de Vita, C. Merlini and P. Zeisser, Energia Nucleare 19 (1972) 676 [26] A. Caridi, F. de Vita, C. Merlini and P. Zeisser, Energia Nucleare 21 (1974) 267 [27] W.R.D. Wilson and J.S. Perrin, Nucl. Sci. Eng. 45 (1971) 99 [28] D. Brucklacher, Physical metailurgy of reactor fuel elements, Proc. Intern. Conf. CEBG Berkeley, Metals Sot., London, 1974, p. 118 [29] J.R. Matthews, Rep. AERE-M 2643 (1974) [30] W. Dienst, J. Nucl. Mater. 61 (1976) 185 [31] A.A. Solomon, Rep. ANL-RDP-13 (1973) 6.4 [32] A.A. Solomon, Rep. ANL-RDP-23 (1973) 5.1