Strength of Materials, VoL 30, No. 2, 1998
R E S I S T A N C E OF A Z I R C O N I U M A L L O Y T O C O R R O S I O N C R A C K I N G U N D E R S T R E S S E S V. N. Kiselevskii, a V. V. Kovalev, a V. A. Stepanenko, b
UDC 539.4
A. I. Stukalov, c and V. M. Gritsina c
We describe a procedure for ultrahigh-frequency thermal treatment of E I lO alloy and present the results of its textural and structural analysis. We also study the influence of the modes of ultrahighfrequency treatment of the zirconium alloy on its long-term strength at a temperature of 653 K in corrosive media. The results of fractographic investigation of the fracture surfaces allow us to determine the mechanisms of initiation and propagation of stress-corrosion cracks in shells made of the alloy under consideration in the intact state and after ultrahigh-frequency treatment.
Introduction. At present, the reorganization of operation of Ukrainian nuclear power plants from the basic service mode to the switching mode seems to be inevitable. Since it is also necessary to increase the depth of fuel depletion, there appears a danger of failures of the shells of fuel elements according to the mechanism of stress-corrosion cracking [ 1,21. At the same time, there exists an efficient procedure for thermal treatment of channel tubes m a d e of I~125 alloy by applying fast high-frequency heating [UHF (ultrahigh-frequency) thermal treatment], which significantly decreases radiation-induced changes in the shape of tubes as a result of the formation of a quasiisotropic structure. The present work is devoted to a study of the influence of fast high-frequency heating of the shells of fuel elements made of t~ll0 alloy on their resistance to stress-corrosion cracking in iodine. It is known that the crystallographic texture is a very important structural parameter which characterizes the resistance of zirconium alloys to stress-corrosion cracking [3]. Since the technology of manufacturing of the shells of fuel elements does not allow one to significantly vary their structure, the tests were carried out on plates with different textures [4]. In this case, it was shown that the influence of textures on the stress-corrosion cracking of zirconium alloys is determined by the orientation of the basal crystallographic (shear) planes relative to external stresses, crystallographic disorientation of grains, and strength of material. We investigate the influence of U H F thermal treatment of the tubes of fuel elements made of t~110 alloy on their resistance to stress-corrosion cracking. The results of the present work demonstrate that U H F thermal treatment is a quite promising method for increasing the resistance of the shells of fuel elements to stress-corrosion cracking. Since the parameters of thermal treatment significantly affect the structure obtained as a result and the mechanism of cracking, in what follows, we determine the optimal modes of U H F thermal treatment and analyze their relationship with the intact structure of the alloy. By varying the parameters of the U H F thermal treatment of tubes of fuel elements, one can induce different types of crystallographic disorientation of grains directly in these tubes but not in the model material. This allows us to investigate the influence of crystallographic textures on the resistance of zirconium alloys to stress-corrosion cracking. For testing, we used tubes O 9.15×7.8×2000 m m in size made of I~ll0 commercial cold-worked alloy in the annealed state. alnstitute for Problems of Strength, National Academy of Sciences of Ukraine, Kiev, Ukraine. bUkrainian National Technical University "Kiev Polytechnical Institute," Kiev, Ukraine. CScientific and Technical Complex "Center of Nuclear Power Engineering" at the Kharkov Physicotechnical Institute, Kharkov, Ukraine. Translated from Problemy Prochnosti, No. 2, pp. 122 - 130, March - April, 1998. Original article submitted September 30, 1997. 0039-2316/98 / 3002--0197520.00 © 1998 Plenum Publishing Corporation
197
I
2
j
4
5
/
7
[--~/
8
5
Fig. I. Block diagram of the "Gorizontal'" installation used for the U H F thermal treatment of the tubes of fuel elements: (1) tube of a fuel element, (2) transporter, (3) guide bushes, (4) sprayer, (5) induction coil, (6) pyrometer, (7) protective housing, (8) inert-gas-supply system. :~.:.'"~..a,~i~:::::::::::::::::::::::::::::::::::::::::::::::: :i::~ .........i: •..:> . ~
....
... ~ : . . ,
:
:!ii-~i
:::;'.'"
:::::::..::::::;~
.:::::..
":~:~i~::.""::i::~~!:.:::$:v.:. ~.":i.,.
.:::~.~:~:.- -.~
"....... ~i!::~:.~!~i~". ...: !.j: :.."
,~..,.:!:~.~.],~: ?~:::. <:
*~?~'": ':!: !i i3 i i ~ ~i !! .~ : i t i .: i i!:iii~iii: ~ ! ~;i~'~!i:i:i:i:i:!:i:i:i:i:?':::*" i : : , . i ~ : :::::::::::::::::::::::
: ....
a
!~i!:;,!:"
::: ! ~::":::
b
~
~'~.::
• ~.
.::::.:
...:~:'-" '
c
Fig. 2. Structures of the original and U H F thermally treated tubes made of I~110 alloy: (a) original state, (b) heating to T = 950°C, (c) heating to T = 1000°C.
T h e r m a l - T r e a t m e n t Procedure. The U H F thermal treatment of tubes was carried out in a "Gorizontal'" installation, which is depicted in Fig. 1. For this purpose, the tubes were translationally moved with a constant velocity through the induction coil of a high-frequency generator and then through a coaxial s p r a y e r with a circular slot. A liquid coolant (water) was sprayed from this slot under pressure. The distance between the induction coil and the cooler was chosen to guarantee the possibility of formation of a uniform distribution of temperature over the cross-section of the tube for the time of passage of a section of the tube heated to a certain temperature along the cooler. T h e heated part of the product was located in a quartz tube filled with argon u n d e r a pressure of 0.04 MPa. The hermetically sealed internal part of the tube was also filled with argon. Argon was used to suppress the process of oxidation of the heated part of the tube. After U H F thermal treatment, the gold-colored oxide film was removed from the surface of the tubes and they were annealed in a vacuum of 1- 10 s m m Hg at a temperature of 833 K for 100 h. The textural characteristics were determined by analyzing the data of X-ray investigations. We used the Morris method for the investigation of materials with close-packed hexagonal lattices by analyzing the intensity of X - r a y reflections in Cu - k a radiation and the weight factors computed for the first 22 lines of reflection. The growth index and the coefficient of linear thermal expansion were computed on a BI~SM computer with the help of special software written in ALGOL-GDR with automatic plotting of inverse pole figures. In this case, the calculations were performed by interpolation of the distribution of the pole density by polynomials of the third degree with b o u n d a r y conditions imposed by the symmetry of the problem. The structure of the tubes was studied with an MIM-8 microscope. In the original state, their structure can be described as a recrystallized matrix of deformed a - Zr alloy with grains 5 - 1 0 / ~ m in size (Fig. 2a). After induction heating to T = 1223 + 25 K followed by quenching in the "Gorizontal'" installation, the alloy consists of equiaxial grains 2 0 - 4 0 / ~ m in size having, for the most part, an acicular structure (Fig. 2b). The procedure of induction heating to a temperature of 1273 + 25 K followed by quenching leads to the formation of an acicular martensitic structure in all grains (Fig. 2c). As temperature increases further, we observe a considerable growth of grains with preservation of the acicular structure. 198
gaff~ ,
oo0~ O.6O
aOg
oo~2 , ~717
~
_
7
2~ 1ot¢ ~ . 5 8 . 4 f
'1.~1
~'~e~~5o
tofo 3.82
O.57
00521.Ol
o002C2.og tor3 k
8
5
1o,,
b
sz-2 ~t3o
I \
t o T o S
t1~°
Iofa~_.........~ 8.02
5
gog2, u7
2.77
ot~z
0.17
~FO
ano
1.27
6 ~o7s ~ tore
ro~,
9 \~
f
2131
/\
.__.z.....-"- t.~ ~ofo0.67 L,
1.20 X
2o~1 rora~a~
~ofo
T
R
Fig. 3. Textures of the original (a) and U H F thermally treated tubes [(b) T = 950°C and (c) T = 1000°C] in the axial (L), tangential (T), and radial (R) directions.
In the original (annealed) state, the texture of the tubes of the fuel elements is characterized by the presence of inverse polar figures in three mutually orthogonal directions (Fig. 3). The poles of the prismatic planes are located along the axis of the tube. T h e degree of unevenness of the texture expressed in terms of the growth index GyL is as high as 0.6-0.7, depending on the conditions of deformation and annealing. The poles of the basal plane are oriented in the tangential (T) and radial (R) directions and, for the growth index, we have GxR = - 0 . 5 4 . After the U H F heating of tubes to T = 1223 + 25 K followed by quenching, we observe an insignificant dispersion of t h e texture in the axial direction (L) and its degree of unevenness in terms of the growth index GxL becomes equal to 0.48. Texture is practically absent in the tubes only after U H F heating to T = 1273 + 25 K followed by quenching in water. In this case, the value of the growth index Gx is close to zero in all directions (Fig. 3c). 199
TABLE 1. Strength of Shells Made of I~110 Alloy in the Tangential Direction at a T e m p e r a t u r e of 653 K Thermal treatment
o0.2, MPa
As-received state M o d e A: fast h e a t i n g to T =
1 2 2 3 K in a r g o n , cooling in water,
MPa
cr u ,
212.0
235.0
226.0
245.0
200.0
210.0
191.0
216.0
237.0
248.0
a n d a n n e a l i n g a t T = 8 3 3 K f o r 5 0 h in a v a c u u m M o d e B: fast h e a l i n g to T =
1273 K in a r g o n , cooling in water,
a n d a n n e a l i n g a t T = 8 3 3 K for 5 0 h in a v a c u u m M o d e C: fast h e a t i n g to T =
1273 K in a r g o n , c o o l i n g in water,
a n d a n n e a l i n g a t T = 8 3 3 K for t 0 0 h in a v a c u u m M o d e D: fast h e a t i n g to T = and annealing atT=
1223 K in a r g o n , cooling in water,
8 3 3 K f o r 10 h in a v a c u u m
crh, MPa
,,/
f~
I
tt7a
I
I
t~ ra
I
I
t I I 1
--...
®
I
I
~
I
101
- ' ~
I
~
I
Ill
!
I
1
I
1
I
I
18z
I I
tf, h
Fig. 4. Dependences of the time to failure on circumferential stresses for the original specimens tested in the absence (1) and in the presence (2) of corrosive media and specimens subjected to thermal treatment in modes A (3), B (4), C (5), and D (6) (see T a b l e l). Experimental Results. T h e tests for short-term strength were carried out at a temperature of 653 K using ring-shaped specimens 2.7 m m in width made of shells in the as-received state and after U H F thermal t r e a t m e n t performed according to the procedure described in [5] (Table i). In testing for short-term strength, the relative residual plasticity of the specimens in the as-received state was as high as 3 0 - 3 5 ~ . A decrease in plasticity by 1 0 - 1 5 ~ , as compared with the as-received state, was observed after thermal treatment. T h e tests for long-term strength (loss of impermeability) of shells in the as-received state and after U H F thermal treatment were carried out at a working temperature of 633 K for a concentration of iodine Ci >- 0.2 m g / c m 2, i.e.,
higher than the critical concentration
at the indicated temperature
(for t~110 alloy at
T = 653 K, we have Ccr > 0.1 m g / c m 2 [3]). T h e curves of long-term strength of specimens subjected to U H F thermal treatment are presented in Fig. 4. For comparison, in Fig. 4, we also display similar curves plotted for the original specimens tested in the absence of corrosive media and after the corresponding thermal treatment. It is easy to see that U H F thermal treatment strongly affects the fracture resistance of the material under long-term loading in corrosive media. The results obtained in the present work corroborate the applicability of the method and equipment proposed in [6]. The method of U H F thermal treatment enables one to attain a substantial increase in the resistance of the material to the development of processes of stress-corrosion cracking. The experimental results can be approximated by the following regression equation: ty = where D and r are experimental constants (Table 2). 200
D
crr,
(1)
TABLE 2. Values of the Coefficients D and r in Eq. (1) Thermal treatment
D
r
As-received state
36.221
14.997
Adequacy variance 0.4828
Mode A
51.206
21.334
0.1040
Mode B
14.595
6.073
0.7080
Mode C
63.675
27.348
0.8761
Mode D
49.652
21.040
0.5618
i)}::::~. . . .
#:
ii ........
..::~g~j{i!ii~:i
iii i! ' iiii!iii :" ======================
a
~!:'
":
": x ~E!.
b
g
:::~'.'.~:.~i~"' ::@.::N::~v..:"~. ~) ":::':':':'?-"
!~::-': "- !ii[ii!..~"" ':::.>:."P" ~ .... ~i
~
::~: .... iii
..::~::.::"~is~:':"..:::::''. ""::':'...:"::::': ' ......... z.. • ......ii-...!!~...
'" .... .
..'i
" .ii~ili~!!~!!::~::T"::~
'
~..~::~:iii!:#::~::ii:-:..... ~:..
!~i~ili~! d
e
f
Fig. 5. Fractographic pictures of the fracture surface of the original specimen: (a) macrofracture ( x 60), (b), (c) point 1 (× 1500 and × 3000, respectively), (d) point 2 (× 1500), (e) point 3 ( x 1500), (f) point 4 (× 1500). After the long-term strength tests, specimens with cracks of various lengths on the outer surface were subjected to fractographic investigations. In the macroscopic fracture surfaces of the original specimens (Fig. 5a), visual inspection performed with small magnifications using optical or electron microscopes reveals zones with different optical characteristics. In the central part, the fracture surface is dark with traces of iridescence. On both sides of this zone, the surface seems to be lustrous and consists of large facets. Closer to the edges, the luster is still present but the facets become smaller. T h e central zone a n d one of the peripheral zones were studied at three points with an electron microscope with great magnification. At point 1, the pattern of the fracture surface is typically intercrystalline with a large n u m b e r of grainboundary cracks, practically around every facet (Fig. 5b). For greater magnifications, at the same point, we observe a decrease in the sharpness of the images of facets but not of the contours of grain-boundary cracks. Most likely, this is explained by corrosion processes proceeding on the fracture surface and leading to the formation either of an oxide film or of an intermetallic compound. In the presence on the surfaces of nonmetallic films, the electron beam interacts with the surface of the specimen in a somewhat different way and, as a result, the image formed by these electrons in the kinescope is not so sharp. It seems likely that the influence of corrosion at point 1 becomes significant not at the onset of crack initiation on the internal surface of a tubular specimen but somewhat later when the crack propagates toward the outer surface. This assumption is corroborated by the fractogram of the surface at point 2 (Fig. 5d). By comparing this fractogram with the fractogram of the surface at point I made with the same 201
....... ",
. , =====================
,'.. ~ i ! ~ i i ~ i.:. i ~........ ~ i : ! :...':' : ! :" .
::. : .
:
:~:..
. ..........: ~.'-:~ .......
"'":'i~
b
a
.....~-:: I===========================. k .../ ~ "....... ..iiiiiiiiiiiiiii~::~iiiii:i:!iii~ . .."i ::~:~iii::i :. -:..- .:..:':~ • -::.: ....:::R:: -" -...:...:.;~.....-..::::::!
::::::::::::::::::::::::::
:...
C
~ I~:#::!::!::!?:.~ i ~ .~!!~-~!~!i .:~i i:i:i'~'i~ !:i~:i::!i:i.! i ~ ! : ' ~ ~i:~:::~::i~ii:~ii!~ii~!i:i!~i!:
..,.::
i!ii~!~i. ~ ::!~i~!!~!i!~!iiz:~i~!~.::~ :~ '~:: -:~ : ::i. "::!:~!
•.~:::....:::..
:.
...~.~.~:..:
":~:!~i::.
..::::::
•
" :~:~:!:!~"
~!!!F
!~ !~;-~:~
i •-"'
:::::::::::::::::::::::::::::::::::::::::::::::::::
::~
N
.......
~ ! ~ .:.# ~i~i!i~is!ir:~.:....':'is:'~::~.."::..":::ii: ~i:~:~:i
"~':"~:~:?'"
:::::
" ~ : :
,
. .:::::
========================================================================================== ::
"
N"" ~
::::::.::: ".: :::.:~..x..~::".'..".~'~:
....:~.~
~:.
"":ii~!!'i
::i::#"::::~:::!:i~
!i:i:~:iii~.-'!~ " : ......... ::iiiiiii:::#:iii~ d
iiii::i?{=:~~~ii !!?ilsii::siii~N!~-".-: e
~ ~ ~ ~ ~ ~ - . ~ f
Fig. 6. Fractographic pictures of the fracture surface of a specimen subjected to UHF treatment: (a), (b) macrofracture (× 60 and × 300, respectively), (c) point I (× 2000), (d) point 2 (× 2000), (e) point 3 (x 2000), (f) point 4 (x 2000), (g) point 4 (x 300). g
magnification, we see that, in the first case, all grain boundary facets are very sharp. In other words, for both fracture surfaces, we have the same basic mechanism of fracture, namely, grain-boundary cracking. At the same time, the influence of corrosion processes at the center of the fracture surface is much weaker, because their duration at this point is smaller. In addition, at both point 1 and point 2 of this zone, in some facets, we observe porous formations in the form of low-density accumulations of small equiaxial pits. At point 3, the fracture process most probably accelerates, i.e., the slow static mode, as in the case of creep, becomes a much faster mode close to dynamic fracture. As a result, the character of macrofracture becomes more complicated. The number of grain boundary facets and surrounding microcracks decreases but, at the same time, the number of medium and large equiaxial pits, ridges of separation, and separate facets covered with numerous multiple-slip lines resembling a fatigue grooved pattern increases (Fig. 5e). In this case, the sharpness of the image is quite high, i.e., the effect of corrosion is negligible. At point 4, we do not observe any symptoms of the effect of corrosion on the fracture process. The micropattern of the fracture surface in this zone is quite complicated. We detect small and large pits, ridges of plastic separation with traces of multiple-slip bands, and elongated plane facets located in steps relative to each other and covered with parallel multiple-slip lines resembling a grooved pattern (Fig. 5f). The macroscopic analysis of the fracture surfaces of specimens subjected to UHF thermal treatment demonstrates that their central part is more dull, porous, dark, and smooth than any other part of the surface (Fig. 6a). It seems likely that the entire area of this zone suffered more or less intense corrosion. The zones lying to the left and to the right of the central part of the fracture surface seem to be formed by lustrous facets. At greater magnifications (the left part of Fig. 6b), these facets are lustrous not because of the realization of a certain 202
mechanism of fracture but as a result of the formation (or precipitation) of intermetallic coatings on these surfaces. Indeed, under the action of electron beams these coatings are charged and, as a consequence, give bright reflections. At points 1 and 2, the patterns of the fracture surface are, for the most part, similar (Figs. 6c,d). In both cases, the fracture surface is not sharp, most likely owing to the formation of a thin foreign coating. The entire surface of the analyzed parts of the fracture surface is covered with chaotically oriented microcracks, either intergranular (it is very difficult to check this) or transcrystalline (less probable). Moreover, some facets contain small equiaxial pits. It appears that some facets are connected by ridges of plastic separation. Separate facets and sometimes even their accumulations are covered with layers of intermetallic compounds, taking the static charge of these facets and giving an intense shining reflection in the image, which almost overshadows the image of the remaining parts of the micropattern. At point 3, the fracture surface contains even more intermetallic compounds that carry static charges and, thus, cause the formation of typical reflections (Fig. 6e). The pattern of the fracture surface is quite complicated and finely dispersed, i.e., we observe pits, porosity formed by smaller, for the most part, equiaxial pits, facets with well-pronounced edges, and microcracks of various directions whose character can hardly be determined. In two or three facets of small length, we see traces of parallel formations similar to short lines which slightly resembl~ fatigue grooves. At point 4, the influence of corrosion is weak. Indeed, the pictures are sharper, the patterns are well-pronounced (Fig. 6f), and the intermetallic accumulations are mainly located in the parts of the fracture surface adjoining the outer and inner surfaces of the cylindrical specimen (Fig. 6g). At greater magnifications, we observe separate medium and large equiaxial pits, plastic bridges between neighboring flat facets covered with parallel linear formations, and a relatively large number of ridges of plastic separation. Special attention should be given to the fact that all lines covering any single plane facet are parallel, whereas the lines of one facet are, generally speaking, not parallel to the lines of another facet and, sometimes, the angle between them can be close to 90 ° . Therefore, if these lines are indeed fatigue grooves, then this means that, in the process of propagation of a fatigue crack, the direction of its front may undergo significant changes (up to perpendicular directions in adjacent segments). This phenomenon can hardly be explained by the correlation of the crystallographic orientations of separate facets with the shape and sizes of an individual grain and its position relative to the neighboring grains. Conclusion. T h e difference between the mechanisms of initiation of stress-corrosion cracks in the original specimens and specimens subjected to UHF treatment is significant. Thus, in original specimens, we observe the realization of an intermetallic mechanism with high degrees of corrosion. At the same time, the mechanism realized in specimens subjected to UHF treatment is of a mixed type because, along with transcrystalline microcracks, one observes fragments of intergranular cracking. In the original specimens, stress-corrosion cracks propagate as a result of intergranular cracking with a significant contribution of corrosion processes. In the UHF-treated specimens, the mechanism of growth of stress-corrosion cracks is mainly transcrystalline. At the same time, the presence of secondary intergranular microcracks of various lengths and orientations corroborates the results of textural investigations concerning the formation of a quasiisotropic structure in UHF-treated shells. REFERENCES
.
F. G. Reshetnikov, Yu. K. Babilashvili, I. S. Golovnin, et al., "Problems of creation of fuel elements for VVt~R-1000 reactors operating at switching nuclear power plants under the conditions of enhanced fuel t
depletion," At. Energ., 64, Issue 4, 258-266 (1988). V. V. Novikov, "Mechanism of iodine cracking of zirconium shells," At. Energ., 71, Issue 1, 33-38 (199!). V. N. Kiselevskii, Strength of Structural Materials of Nuclear Reactors [in Russianl, Naukova Dumka, Kiev (1990). V. F. Zelenskii, A. I. Stukalov, I. M. Neklyudov, et al., "Fast high-frequency thermal treatment and radiation-induced growth of zirconium alloys," in: Proceedings of the Internat. Conf. on the Radiation Materials Science (Mushta, May 1990) [in Russian], Vol. 4, Mushta (1990), pp. 55-63. V. A. Borisenko, A. I. Gurarii, and V. P. Krashchenko, A Method for the Investigation of Mechanical Properties of Materials of Metallic Ring-Shaped Specimens at High Temperatures [in Russian], Preprint, Institute for Problems of Strength, Academy of Sciences of Ukraine, Kiev (1986). V. N. Kiselevskii, V. V. Kovalev, and G. S. Rossikhin, "Kinetics of fracture processes in the shells of fuel elements made of zirconium alloys in iodine," ProbL Prochn., No. 9, 30--39 ( 1 9 9 4 ) . r
.
3. .
.
.
203