Studies on the sintering behaviour of uranium dioxide powder compacts

Journal of Nuclear Materials North-Holland, Amsterdam

261

158 (1988) 261-266

STUDIES ON THE SINTERING BEHAVIOUR POWDER COMPACTS

OF URANIUM

DIOXIDE

Pranab DAS * and Ranjit CHOWDHURY Research and Development Center for Iron and Steel, Doranda, Ranchi 834002, India Received

16 April 1987; accepted

11 March

1988

Uranium dioxide fuel pellets are normally made from their precursor ammo~um diuranate, followed by calcination, subsequent reduction to sinterable grade powders and a post operation treatment of pressing and sintering. The low temperature calcined powders, usually exhibiting non-crystalline behaviour (under X-ray diffraction studies) progressively transforms into a crystalline variety on subsequent heat treatment at higher temperature. It is observed however that powders calcined between 800 to 9oO°C exhibit enhanced densification behaviour when sintered at higher temperatures. The isothermal shrinkage versus time plot of the sintered compacts are well described by a hyperbolic relationship which takes care of the observed shrinkage (h) as caused due to a cumulative effect from the initial sintering of the powder compacts at zero time (a) and that caused due to the structural transformation from a non-crystalline modification with increased thermal treatment (p). The derived equation is a modification of the sintering mechanism of the viscous flow type proposed by Frenkel, involving sintering of an amorphous phase, the viscosity of the latter is presumed to increase with increasing thermal treatment to assume the final modified form as X=r/(a+j3t), where t = time, h = shrinkage

and LYand & are the unknown

1. Introduction Uranium dioxide fuel pellets for nuclear power reactors are normally made by cold pressing and sintering of the calcined and reduced ammonium di-uranate powder compacts. The sintering characteristics of the pellets as dependent upon their precursor powders calcined at various temperatures have been a subject of intensive study for the last thirty years or more. Usually it is observed that when subjected to increasing thermal treatment, the nature of the calcined powders change from a poor to a well developed crystalline modification as revealed from X-ray diffraction studies. Clearly such a phenomenon takes into account a physico-chemical change involved and during sintering of such powders expression relating shrinkage/ densification to the established mechanisms of sintering needs modification. While the role of ‘activity’ of the calcined powders has been held responsible in many earlier investigations * Present address: Bharat Refratories, Ltd., C/O RDCIS, Ran&i, India.

~22-3115/88/$03.50 (North-Holland

parameters.

to account for the apparently anomalous densification behaviour of the powder compacts during the sintering a clear understanding on the subject appears scanty. The present work is an effort to explain this apparent anomaly at least in a semi-quantitative way taking help of the established sintering mechanisms and models. The explanation lend its support on simmilar observations made earlier on materials exhibiting phase transformation during sintering [1,2].

2. Theory The simplest sintering mechanism applicable to those noncrystalline low temperature calcined powders is very similar to the one proposed by Frenkel [3] whose mathematical expression in terms of shrinkage is X = Kot/qr,

0 Elsevier Science Publishers B.V.

Physics

Publishing

Division)

(1)

where A is the relative shrinkage, t is the time, r the grain size, ?I the apparent viscosity of the material (reciprocal of plasticity), u the surface tension of the solid and K is the proportionality constant. Due to structural modification of the powder with increasing

P. f)as, R. Chowdh~~ / Sintering beh~viour of U#, powder compucrs

262

thermal treatment the sintering ability of the material will diminish progressively as if its non-crystalline character diminishes i.e. as if the apparent viscosity of the material is increasing following the equation rl=rlo(l

+ct),

(2)

where q0 is initial apparent viscosity, t is the rate of increase of viscosity and t is the time. It seems reasonable to presume that in the first stage the structural modification will affect 17, while the driving force for sintering (surface tension) remains constant such that the apparent viscosity of the material increases with time. Such a hypothesis relate viscosity with time linearily as shown in eq. (2). Substituting eq. (2) in eq. (l), one obtains X=1/(170Y(r)/aK+I70Y(r)Et/crK).

(3)

This expression establishes a hyperbolic relationship between shrinkage and time which is of the form X=t/(LYf&),

(4)

where a = ~,y( r)/uK and p = ~y( r)r/aK. The experimental plot of X-‘l versus time will give a linear relationship and the significance of the parameters (Y the initial shrinkage rate and p are (Y-I = (dhfdt),,,, and /3-’ = lim X,,, = h, which is the limit of shrinkage at infinite time. (P/a) has the expression of rate of increase of viscosity (from eq. (2)) and may be defined as initial shrinkage rate relative to the shrinkage limit (a-‘//l-‘) or the relative shrinkage rate. Further it may be derived that

involved. Since isothermal conditions of sintering is not instantaneously reached but a certain amount of contraction takes place before actual measurements are made, the origin of the isothermal curve can not be found experimentally. As a result regression analysis of the data points were extraplotted to zero time from which the initial shrinkage rate, a-‘, was found out.

3. Experimental Urania powders for this study were prepared from calcination and reduction of a~onium diuranate powders at temperatures varying between 500-10000 * C in hydrogen atmosphere for three hours. The physical properties of the powders are shown in table 1. Pellets (approximately 10.0 mm diameter x 3.0 mm height) were prepared by cold pressing of the powders at 20.0 T.S.1 (T.S.1: tons per square inch). The green density of each individual pellet prepared from these different temperatures calcined powders were measured from direct geometry measurements. A variation between 54.0 to 55.0% T.D. (TD: theoretical density) was observed for all the different compacts studied. Isothermal sintering were carried out at temperatures varying between 1000 to 1400° C for various soaking periods limited to a maximum time of 1000 min. Bulk density of the pellets were determined by water displacement method from which volume shrinkage of each individual pellet was calculated. h was found out from the expression X= AV/3V,, where AV is the volume shrinkage and V, is the original volume of the pellet under measurement,

4. Results or Pia

= [log

~lo”/l~l,m.

(5)

This expression (on the average) gives a relationship between rate of increase fo viscosity or relative shrinkage rate which has a logarithmic (0 c t -=z00) relationship with the observed shrinkage. Considering further that a is related to shrinkage rate at t = 0, or the actual sintering behaviour at zero time while fl is related to the shrinkage behaviour associated with the physico-chemical change of the powders (t + co), and activation energy plot for the sintering process can be made between In(tY-‘) versus l/T (T is the absolute temperat~e) where the slope will determine the apparent activation energy of the process during actual sintering. without having any bias on the associated physico-chemical property changes

In fig. 1 a plot of X-‘r versus t and h versus t for various compacts made from powders calcined at 500 o C

Table 1 Physical characteristics of the uranium dioxide powders Calcination temperature (“C) 500 6W 800 900 1000

Specific surface area tm’/gf 6.0 6.2 3.5 2.6 1.1

Crystallite size

Powder density

(A)

&m/cc)

154.0 _

9.7 10.0 10.3 10.5 10.8

250.0 315.0

P. Das, R. Chowdhuty

/ Sintering behaviour of UO, powder compacts

h”st

AND i’t

VS t

CALCINED AT

/

PLOT

FOR

POWDERS

263 500% SMITERED

5 ii I2500 5 u . Y

2000

&a f e-1500 TX

Fig. 1. Plot of A-‘r versus time for uranium dioxide powders calcined and sintered at various temperatures.

and sintered in the temperature range lOOO-1200°C where found to exhibit a good linear relationship (in the former case). The correlation coefficient was found bet-

500 600

ter than 0.96 for all the different experimental samples. To predict a and @ values more precisely, regression analysis and least square fitting of the data points were made. Computed from fig. 1, the absolute values of 100/a were 6.07 X 10W3, 0.07 and 0.2 and the absolute

“C CALCINATMN % CALCINATIO N

700%

CALCINATK)N

8OO’C

CALCINATION

900%

CALCINATION

CALCtNATION TEMP. 0 soo% e 600% e 700% 0 e

fJoo”c 900°C

10 I 1ooot SINTERING

TEMPERATURE

(‘C)

Fig. 2. Plot of (100/a) versus sintering temperature.

I 12oooc

rroo’t: SINTERING

I

I

1300°C

TEMPERATURE

1400% f “C,

Fig. 3. Plot of (lOO,$) versus sintering temperature.

P. Das, R. Chowdb~~ / Sinfering behaviou~ of UO, powder compacts

264

SINTERING

values of lOO/& were 26.9, 37.9 and 46.2 at the sintered temperatures 1000 “C, 1100 o C and 1200 o C, respectively. In fig. 2 a plot between 100/a and various sintering temperatures is shown for powders calcined between 500 and 900°C. It is interesting to observe that the X00 ’ C calcined powders show a high cy-’ value. In fig. 3 a plot of 100/p versus sintering temperature is made. It will be observed that powders show their characteristic differences at a lower sintering temperature, nameiy 1000°C. The powder calcined at 500’ C shows the highest fi-’ value which diminishes progresively with increasing calcination temperature and the 900 o C calcined powder exhibits the lowest fi-’ value. Such observed differences, however, become marginal with increasing sintering temperature and at 14OO’C the powders exhibit a negligible difference amongst themselves. In fig. 4 a plot between (p/a) and the sintering temperature for each individual calcined powder is shown. A sharp increase for the 800 o C calcined powder is observed here again conforming the observations made in fig. 2. As (/3/a) indicates the relative shrinkage rate this is highest for the 800 o C calcined powder. It is

-1200

TEMI?

“c

--~1300°C

CALCINATION

TEMPERATURE

t°C 1

Fig. 5. Plot of (lOO/rr) versus caicination temperature. CALCINATION

TEM!?

0

500

Oc

8

600

“c

SINTERING

(B 7oo”c I

1000

1200 .

SINTERING

Fig. 4. Plot of (P/a)

1000 “c 1100”c

-.-

1400”c

1400

1300 TEMPERATURE

TEMPERATURE

---

{‘Cl

versus sintering temperature.

CALCINATION

TEMPERATURE

f°Cf

Fig. 6. Plot of (100/,8) versus calcination temperature.

P. Das, R. Chowdhwy

CALCINATION

/ Sintering behaviour of UO, powder compacts

TEMPERATSE 0 500°C e 600°C .

7oo”c

a” i 8 0

e c

8e

I

5.0

I

I

I

60

7.0

6.0

Ln(llT)X104

Fig. 7. Plot of ln(lOO/a) versus ln(l/T) for determination of activation energy of the sintering Process.

concluded further that such a difference arises out of a high initial shrinkage rate (a-‘) which is seen in fig. 2. It is observed further from fig. 5 that this significant change in a- ’ for the 800 o C calcined powder becomes more pronounced with increasing sintering temperature. A simmilar plot (fig. 6) as regards change in p-i values with respect to the sintering temperature indicates that the characteristic difference between the powders loose significance at higher sintering temperatures, as the values exhibit marginal difference. In fig. 7 an activation energy plot for the sintering process (In a-i versus ln(l/T)), made for powders calcined between 500 to 700°C, by the least square fitting method yields a value of 49.2 kcal/mol.

5. Discussion

Experimental results conclude that the initial shrinkage behaviour of compacts made out of 800 o C calcined powders are significantly different from others. The

265

limit of shrinkage at infinite time show marginal variation with change in calcination treatment. Any observed difference in the relative shrinkage rate (p/a) is a result arising out of variation in the initial shrinkage rate of the powders. Observations mentioned as above are a result of two simultaneously occurring phenomenon, namely the sintering of the particles at the experimental temperatures mentioned and a simultaneous process of rearrangement of the atoms within the crystalline lattice leading to an ordered structure and its associated role in influecing the sintering phenomenon. While the former is purely physical in nature based on surface energy considerations, the later is aided with an associated chemical or physico-chemical process and its effect in influencing the surface energy of the system influences the observed shrinkage behaviour. The phenomenon continues till a minimum free energy configuration is reached in the system. It is, however, difficult to delineate each process separately, as the individual kinetic processes will have overlapping influence with test temperature and time. It will be observed moreover that the physical significance of the equation relates to an expression which states that at any instant the total time realised to achieve unit shrinkage is a function of (1) time independent (zero time) shrinkage, a, and (2) a time dependent factor, p. While a takes care of the physical phenomenon namely sintering, the /3 factor takes care of the associated changes in the material characteristics. In the present example, this is a rearrangement process to lead to an ordered structure from a disordered lattice. The change in p values will be more pronounced in examples where the initial rearrangement process is slow and will exhibit marginal difference with time and temperature. The nature of the curves in fig. 3 examplifies this. It may be concluded, however, that in the present example the abnormal rise in a-' values with time for the 800° C calcined powders relate to a phenomenon where the two processes superimpose with each other. It is known, however, that such changes are always associated with a non-equilibrium phase at the transformation temperature and while equilibrium between the two dissimilar crystallographic modifications tend to occur at the interface the lattice is strained giving rise to the well known ‘ transformation plasticity’ phenomenon. Anomalous diffusion behaviour and increased mobility of the atoms leads to an enhanced densification behaviour. A significant rise in a-’ value perhaps is a consequence of that effect. Similar behaviour has been reported in case of polymorphic phase transformation in other related material systems also.

266

P. Das, R. ~howdb~~

/ Sinteri’ng behavjaur

6. Conclusion Sintering behaviour of uranium dioxide powder eompacts prepared from 500 to lOOO*C calcined powders and sintered in the temperature range of 1000 to 1400 o C can be well described by a hyperbolic relationship between shrinkage and time. The predicted equation takes care of the sintering phenomenon due primarily to the physical sintering process (a) and that caused with any change in the material characteristic resulting from physico-chemical changes in the process (p) involved. In the present example the enhanced densification of the powder compacts at an intermediate Cal&nation temperature can be well described on the ‘ transformation plasticity’ phenomenon. The observed behaviour falls in line with similar observed phenomena in other material systems also [4,5].

Acknowledgements The authors wish to thank the staff members of the metallurgy division. B.A.R.C., Bombay for their kind

of UO, powder compacts

help in different stages of the experimental work. The authors express their sincere thanks to the management of B.R.L., B.S. City for kind help and permission for publication, The help rendered by Shri S. Dutta of the RDCIS Sub-Centre, IISCO, Burnpur in various secretarial work is gratefully acknowledged.

References

[l] V.K. Murthy and S.V.K. Rao, Trans. Ind. Ceram. Sot. 27 (1968) 16. [2] A.C.D. Chaklader, Proc. J. Brit. Ceram. Sot. 15 (1970) 216. [3] J. Frenkel, J. Phys. (USSR) 9 (1945) 385. [4] A.C.D. Chaklader, Bull. Am. Ceram. Sot. 54 (1975) 399. [5] A.C.D. Chaklader, Deformation of Ceramic Materials. Material Science Research Series, Eds. R.C. Bradth and R.E. Tresler (Plenum Press, New York, 1975) p. 425.