- Email: [email protected]
Fabrication and measurement of hoop strength of SiC triplex tube for nuclear fuel cladding applications
Journal of Nuclear Materials 458 (2015) 29–36
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Fabrication and measurement of hoop strength of SiC triplex tube for nuclear fuel cladding applications Daejong Kim ⇑, Hyun-Geun Lee, Ji Yeon Park, Weon-Ju Kim Nuclear Materials Division, Korea Atomic Energy Research Institute, 989-111 Daedeok-daero, Yuseong-gu, Daejeon 305-353, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 June 2014 Accepted 26 November 2014 Available online 3 December 2014
a b s t r a c t The SiC ceramics are under investigation for the fuel cladding in the light water nuclear reactors because of its excellent high temperature strength and corrosion resistance against hot steam under the severe accident conditions. In this study, the SiC triplex tubes consisting of a SiC inner layer, a SiC/PyC/SiC intermediate layer, and a SiC outer layer were fabricated by the chemical vapor processes. The hoop strength and fracture behaviors of the SiC triplex tube were investigated. The SiC triplex tubes fabricated at the high ratio of H2/MTS had a quite high average strength with a relatively small standard deviation. The hoop strength of the composite tubes tends to increase with the volume fraction of the reinforced fibers. The highest fiber volume fraction was obtained using Tyranno SA3-0.8k with the dense winding patterns such as bamboo-like mosaic pattern, which resulted in the high hoop strength compared to other fibers of Tyranno SA3-1.6k and Hi-Nicalon Type S. Hoop strength also increased slightly as the winding angle increased from 45° to 65°. Fracture behaviors of the SiC triplex tube were investigated via the observation of microstructure of the failed samples. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction The current light water reactors (LWRs) have used Zr alloys as a nuclear fuel cladding due to its superior properties such as a low neutron absorption cross-section and a good neutron irradiation resistance under the operation conditions of the reactor. However, Zr alloys react rapidly with hot steam produced during severe accidents such as loss-of-coolant-accidents (LOCA), undergoing significant recession and degradation of mechanical strength, and producing massive hydrogen [1]. Therefore, there have been extensive efforts to develop accident tolerance fuel cladding materials to replace Zr alloys. There are several new design concepts of LWR fuel cladding, categorized by Zr-based concepts such as SiC or MAX phase coated Zr alloy, SiC composite + Zr alloy or Zr–Mo–Zr sandwich hybrid, and fully replacement concepts such as Fe–Cr–Al alloys, Mo alloys, SiC triplex, and stainless steels [2]. Most of the new concepts have considered a use of SiC ceramics as their constituent because the hydrogen liberation rate of SiC is hundreds times less than Zr alloys [1,3]. Furthermore, SiC and its composites have excellent high temperature mechanical properties, a low neutron absorption cross section, a high melting point, a good neutron irradiation resistance, and little irradiation creep, compared to Zr alloys [4–6]. ⇑ Corresponding author. Tel.: +82 42 868 4559; fax: +82 42 868 8549. E-mail address: [email protected] (D. Kim). http://dx.doi.org/10.1016/j.jnucmat.2014.11.117 0022-3115/Ó 2014 Elsevier B.V. All rights reserved.
Among various nuclear fuel concepts, the SiC triplex proposed by CTP Ltd. has been considered as one of the leading concepts because of its larger safety margins under beyond-design basis severe accident and high burnup capability [7,8]. However, there are still many technical issues such as understanding of statistical failure of the SiC triplex tubes, stress concentration on an inner SiC layer and failure due to swelling-induced stress, fission products release due to microcracking in a matrix SiC phase, chemical compatibility between coolant and cladding under operating conditions, development of the hermetic joining method of end-plug, and manufacturability of long cladding tubes and manufacturing defects [9–13]. The SiC triplex fuel cladding consists of a monolith SiC inner layer, a SiC/SiC composite intermediate layer, and a monolith SiC outer layer, as shown in Fig. 1. A primary function of the SiC inner layer is to prevent radioactive fission products from escaping the fuel into the coolant, the SiC/SiC composite layer enhances the strength of the cladding tube and provides better fracture toughness, and the SiC outer layer protects a composite layer against corrosive coolant water at high temperatures. In order to exhibit the best performance during neutron irradiation, each layer should be pure, well-crystallized and near-stoichiometric [14,15]. With regard to fiber selection only Tyranno SA3 and Hi-Nicalon Type S SiC fibers meet the properties required among various commercial products. In case of monolith SiC and SiC matrix phases, high quality SiC phases is usually obtained by
30
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
the chemical vapor processes. It has been recently reported that the chemical vapor infiltration (CVI)- and the nano-infiltration and transient eutectic phase (NITE)-processed SiC composites have best neutron irradiation resistance [16–18]. The composite produced by other process methods such as polymer impregnation and pyrolysis (PIP) and liquid silicon infiltration (LSI) contains impurities which would result in the poor performance during irradiation such as local swelling and strength degradation [19,20]. Therefore, in this study, the SiC triplex tubes which consisted of a monolith SiC inner layer, a SiC/PyC/SiC composite intermediate layer and a monolith SiC outer layer were fabricated by the chemical vapor processes such as CVD and CVI. Influences of filament winding methods, type of SiC fibers on hoop strength of the triplex tubes were investigated. The damage process during the hoop tests were examined via microstructure observation. 2. Experimental procedure 2.1. Fabrication of SiC triplex composite tube A monolith SiC inner layer was uniformly deposited with about 320.3 ± 6.4 lm by a chemical vapor deposition method using methyltricholorosilane (MTS, CH3SiCl3) onto the high purity graphite rods with 8.5 mm in diameter and a length of 100 mm. Then the SiC fibers were applied to a SiC-coated cylindrical mandrel by a filament winding method [21]. SiC fiber bundles (tows) consisting of 0.8–1.6k fibers were provided on a spool under a little tension which was generated only by the friction of the fiber supplying system. 1–2 helical plies of SiC fiber bundles covered the mandrel. The properties of SiC fibers are listed in Table 1. The commercial generation III SiC fibers such as Tyranno SA3 and Hi-Nicalon Type S were used in this study. They contain a small amount of impurities and they are crystallized and near-stoichiometric, so called nuclear grade SiC fibers. A fiber volume fraction was controlled by changing the band width of each fiber. For helical winding it is usual to lay fibers side by side, to give complete coverage as the winding progresses. Because of the difference in band width, the reciprocating motions of a transverse carriage for the complete coverage were varied from 12 to 21 times, which results in the change in fiber volume fraction and corresponding thickness of the composite intermediate layer in the range of 240–430 lm. As the input value of the band width is smaller than the actual value, accordingly the pattern is open which results in the less fiber volume fraction. The winding pattern was controlled by adjusting the position of second fiber trajectory. Fig. 2 shows four phase winding patterns wound with the winding angle of ±55° using Tyranno SA3-0.8k and schematic illustrations. Winding angles were varied to ±45°, ±55°, and ±65° using Tyranno SA3-0.8k with the 1-return winding pattern. After the filament winding, PyC was deposited as an interphase of the SiCf/SiC composite onto SiC fibers by dehydrogenation of CH4
at 1100 °C using by a CVD method. For the infiltration of SiC matrix phase, the isothermal CVI was processed at 1000 °C at the low pressure of 3.3 kPa to reduce density gradient to the radial direction of the tubular specimens. Finally a SiC outer layer was deposited with the same conditions to the matrix infiltration in sequent. The SiC outer layer exhibited the wide thick variation of 39.8 ± 18.6 lm. 2.2. Measurement of hoop strength The tubular specimens for hoop strength measurements were prepared by following procedure. The substrate graphite rod was burned out at 1200 °C for 30 min and 900 °C for 21 h in the atmospheric environment. Graphite ash remained on the inner surface of the composite tube was removed by air blowing and ultrasonic cleaning. The tubular specimens have a dimension of a length of 10 and 30 mm, an inner diameter of 8.5 mm, and an outer diameter of 9.6–10.1 mm. Then the polyurethane plug with 8.45 mm in diameter and a length of 8 and 22 mm was inserted into the tubular specimen with a length of 10 mm and 30 mm, respectively. Hoop strength was measured via internal pressurization at room temperature using screw driven universal testing machine (Instron 4465, load cell capacity = 5 kN). Axial pressure was applied with loading rate of 0.01 mm/s to the cylindrical polyurethane plug. A series of hoop tests were carried out at room temperature and each test was repeated 2–7 times. Weibull modulus of the hoop strength was obtained using 20 samples with 685 ± 41 lm thickness and a length of 10 mm. Radial displacement measurements were made by attaching four displacement transducers (Kyowa, Co. Ltd) to the central section of the outer tube surface. Hoop strength of tubular SiC composite specimens, rh was calculated by [22]
rh ¼
r2i P r 2o 1 þ r 2i r 2o r 2i
ð1Þ
where P is the internal pressure, ri and ro are the inner and outer radii of the tubular specimens, respectively. Contact strength was typically 50–100 N which was determined at the increasing point of radial displacement. 3. Results and discussion 3.1. Microstructure of SiC triplex tube Fig. 3 shows microstructures of a SiC triplex tube and its constituent layers of a SiC inner layer and a Tyranno SA3-reinforced SiCf/ SiC composite layer. The CVD b-SiC inner layer has columnar structure elongated to the growth direction, as shown in Fig. 3(b). The layer is dense without voids. The SiCf/SiC composite layer consists of SiC fibers, PyC interphase, and SiC matrix, as shown in Fig. 3(c). The PyC interphase of about 200 nm in thickness was deposited onto the SiC fibers. The intrabundle space within the fiber woven composite layer was densely filled with SiC by a CVI process. Some circumferential voids were formed in the SiC matrix but the fraction of voids is a quite low compared to typical CVI-processed SiCf/SiC composite [23]. The columnar grains of the SiC matrix are grown to the radial direction of the fibers. The grain size is much smaller than the SiC inner layer because SiC matrix was deposited at the lower temperature of 1000 °C. 3.2. Hoop strength of SiC triplex tube
Fig. 1. SiC triplex tube concept for LWR fuel cladding application.
3.2.1. Hoop strength and Weibull modulus Fig. 4 shows the Weibull statistical plot of ultimate hoop strength of SiC triplex tubes whose intermediate SiCf/SiC composite layer were infiltrated at 1000 °C at H2/MTS ratio of 20. The
31
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36 Table 1 Properties of SiC fibers used in this study. SiC fiber
Compositions (wt%)
Filament dia. (mm)
Filaments/yarn
Tensile strength (GPa)
Tensile modulus (GPa)
Ply no.
Tyranno SA3-0.8k Tyranno SA3-1.6k Hi-Nicalon Type S
68Si + 32C + 0.6Al 68Si + 32C + 0.6Al 69Si + 31C + 0.2O
7.5 7.5 12
800 1600 500
2.1 2.1 2.6
395 395 420
2 1 1
Fig. 2. (a) Four phase winding patterns and (b) schematic illustrations wound by a helical filament winding method using Tyranno SA3-0.8k with the winding angle of ±55°.
Fig. 3. (a) SEM microstructures of the SiC triplex tube and back scattered electron images showing microstructures of (a) a monolith SiC inner layer and (b) a Tyranno SA3 reinforced SiCf/SiC composite layer.
ultimate hoop strength for 20 samples was within the 235.0– 337.8 MPa. The SiC triplex tubes had a quite high average strength, 282.4 MPa, with a relatively small standard deviation, 44.3 MPa. It
has been known that CVD SiC has Weibull modulus of 5–11 depending on the volume of specimens and test methods. 2D plane-woven Tyranno SA or Hi Nicalon SiC fiber reinforced SiCf/
32
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
Fig. 4. Weibull statistical plot of ultimate hoop strength of SiC triplex tubes.
SiC composite has Weibull modulus of about 10 [17,6,24,25]. In case of filament wound composite tubes, the matrix density of the SiCf/SiC composite layer was higher than that of the 2D plane-woven composite which was measured to be about 2.92 g/ cm3. Therefore, Weibull modulus of the SiC triplex tubes was high to be 11.05. This value is comparable or higher than CVD SiC and plane-woven SiCf/SiC composite. Since all samples for Weibull statistics were taken from the same batch process, however, the deviation of strength can increase due to the difference in matrix density between batch processes. When considering that an orthotropic lamina composite with its principal material (L and T) axes oriented at an angle, h with the reference coordinate axes (x and y), stresses can be transformed from one set of axes to another [26,27]:
8 9 > < rL > =
9 38 > < rx > = 7 rT ¼ 6 4 sin2 h cos2 h 2 sinh cosh 5 ry > > > : > sLT ; sinh cosh sinhcos h cos2 h sin2 h : sxy ; 2
cos2 h
2
sin h
2sinhcos h
ð2Þ
If a stress, rx is applied to an orthotropic lamina, the longitudinal, shear, and transverse stress of the composite along the L and T directions can be calculated by the stress-transformation law:
rL ¼ rx cos2 h sLT ¼ rx sin h cos h rT ¼ rx sin2 h
ð3Þ ð4Þ ð5Þ
Based on the maximum stress theory associated with a weakest link phenomenon, the failure mechanism changes from tensile fraction of the fibers to matrix/interphase shear to matrix/interphase cleavage as the fiber angle increases from 0° to 90°. When a unidirectional composite which is loaded parallel to the fiber direction, with angled-ply orientation, and perpendicular to the fiber direction, the fiber volume fraction, Vf ideally has an effect on the normal and shearing stress as follows [27]:
rL ¼ rf V f þ rm ð1 V f Þ s s sLT ¼ GLT f V f þ m ð1 V f Þ Gf
rT ¼ rm
Gm
ET ð1 V 1=3 Þ Em
Assuming the uniaxial tensile stress is applied to the circumferential direction of the SiC triplex tube during the expanding plug test, the hoop strength of the SiC triplex tube with the fixed winding angle increased with the aligned fiber volume fraction, as shown in Fig. 5. All samples have almost same thickness of SiC inner with about 320 lm and outer layers with 40 lm. An increase in hoop strength could be contributed by a SiCf/SiC composite layer whose thickness increased from 250 to 420 lm with the fiber volume fraction. Therefore, it is important for obtaining higher hoop strength to increase the fiber volume fraction per unit composite volume. The highest fiber volume fraction was obtained when the composite layer was wound by Tyranno SA3-0.8k fiber which resulted in the highest hoop strength. In the case of the Hi-Nicalon Type S-reinforced triplex tubes, the fiber volume fraction is less than Tyranno SA3 that has the same winding architecture because a diameter of Hi-Nicalon Type S is higher than Tyranno SA3 and thus Hi-Nicalon Type S is less flexible. Although Hi-Nicalon Type S has higher tensile strength and tensile modulus (Table 1), its composite tube has similar hoop strength with the Tyranno SA3-reinforced triplex tube. 3.2.2. Effect of filament winding pattern and angle The winding pattern itself leads to the different strength and strain field [28]. Moreover, it has a significant effect on fiber volume fraction and void distribution between fiber bundles [29]. As shown in Fig. 2, the fiber bundle for the bamboo-like mosaic pattern such as 1-return and 2-return lays side by side and there is little interference between fiber bundles with ±h angles. Therefore, the void fraction of a SiCf/SiC composite layer after the CVI process was relatively low which was 10.3% for 1-return and 11.1% for 2-return. A small size of voids is distributed mainly at the intrabundle site of the fiber. Relatively large voids exist in the interbundle and the inter-ply region, as shown in Fig. 6(a) and (b). For the triplex tubes with 6-return and 9-return winding patterns, on the other hand, two or three times space of the band width exists between fiber bundles with +h angle. Fiber bundles with h angle are slightly lifted due to interference of +h angle fiber bundle, resulting in the formation of large interbundle voids, which results in an increase in thickness of a composite layer. As shown in Fig. 6(c) and (d), the fiber volume fraction is relatively low and large interbundle and inter-ply voids are formed after the CVI process. The void fractions of the composite layer were relatively high which were 16.8% and 14.9% for the composite layers with 6-return and 9-return, respectively. These large circumferential voids may be detrimental on the integrity of fuel cladding. When a SiC inner layer is failed, fission products can easily release through the microcracks in matrix phase and voids into the
ð6Þ ð7Þ ð8Þ
where rf and rm are the ultimate strength of the fiber and the matrix stress at the fiber fracture strain, respectively. E and G are elastic and shear modulus, respectively. As the fiber volume fraction, Vf increases, therefore the fibers carry more of the applied load at a low angle to the fiber direction and the strength of the composite increases with the aligned fiber volume fraction.
Fig. 5. Hoop strength of the SiC triplex tubes as a function of the fiber volume fraction.
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
33
Fig. 6. SEM microstructure of the SiC triplex tubes reinforced by Tyranno SA3-0.8k with the winding patterns of (a) 1-return, (b) 2-return, (c) 6-return, and (b) 9-return.
coolant. Furthermore, thermal conductivity of the SiC triplex tube will be decreased. Fig. 7 shows the hoop strength of the SiC triplex tubes with various winding patterns and winding angles which were reinforced by Tyranno SA3-0.8k. The hoop strength tends to increase with increasing the fiber volume fraction depending on the winding patterns, as shown in Fig. 7(a). However, the hoop strength of the tubes with 2, 6, 9-return winding patterns is higher than that of 1-return although the tube with 1-return winding pattern had highest fiber volume fraction. It is believed due to an increase in the fraction of the composite layer in the whole triplex tube. Fig. 7(b) shows the hoop strength of the triplex tubes reinforced by Tyranno SA3-0.8k with the different winding angle. Since fibers have superior mechanical properties along their lengths, higher orientations of the fibers provide the higher hoop strength of the composite tube [30,31]. Therefore, the hoop strength of the tube increases as the winding angle gets higher. However, the effect of winding angle on the hoop strength is less pronounced on the corresponding properties of 45° and 65° wound tubes, compared to conventional SiCf/SiC composites which could be caused by that the fraction of the fiber reinforced composite layer is not much in the triplex tube. 3.3. Fracture behavior Fig. 8(a) shows the applied load-axial displacement curves of the SiC triplex tubes. Load gradually increases until the polyurethane plug contacts with the tubular specimen at the initial stage, and it steeply increased after the contact. Especially some load drops are observed before failure of the tube which might be caused by the formation of the large axial crack. The stress–strain curve displays an initial almost linear up to almost a half of the curve, as shown in Fig. 8(b). The onset of non-linearity could be associated with matrix cracking and manifests itself in degradation
Fig. 7. Hoop strength of the SiC triplex tubes depending on (a) winding pattern and (b) winding angle.
34
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
(a)
(b) Fig. 8. (a) Typical load–axial displacement and (b) stress–radial displacement curves of the SiC triplex tubes.
in tube stiffness after a significant load drop and an increase in radial displacement. Fig. 9 shows the microstructure of the triplex tube interrupted right after a first load drop during the hoop test. Failure of the SiC inner layer obviously takes place, as shown in Fig. 9(a). However, the crack from the inner layer does not propagated immediately into the composite layer. In the SiC triplex tube, PyC with about 200 nm in thickness exists between the SiC inner and composite layers, which was formed during the deposition of PyC interphase on SiC fibers. When the crack propagates from inner layer to composite layer, therefore the PyC interphase become the preferred site for crack propagation. It indicates that the existence of a thin PyC layer play a role of an obstacle of crack propagation. However, after a failure of the SiC inner layer, the stress concentrated on a SiCf/SiC composite layer, followed by a large expansion of the quasi-ductile composite layer. Therefore, a number of intrabundle microcracks were developed primarily along the matrix/ fiber interface to the circumferential direction by shear stress at the initial fracture stage. A small population of the radial cracks was also observed, as shown in Fig. 9(c) and (d). Fig. 10 shows load-axial displacement curves of SiC triplex tubes with 10 mm length and Weibull distribution of the hoop strength at the initial load drop associated with a fracture of a SiC inner layer. Most of the triplex tube samples experienced load drop phenomena except for 3 samples. Hoop strength at the first load drop was in the wide range of 155.3–309.5 MPa. The average hoop strength was 226.95 MPa. Although these values may not represent the crack initiation, they imply the onset of a significant axial crack in the SiC inner layer. Weibull modulus for hoop strength at the first load drop was calculated to be 5.30 which is much lower than ultimate hoop strength. Failure of a SiC inner layer at the lower stress and the wide distribution of failure strength may have a detrimental effect on the reliability of the
Fig. 9. Microstructure of the SiC triplex tube after a first load drop: (a) a crack in a SiC inner layer and (b) a matrix crack in a SiCf/SiC composite layer at low magnification and (c) high magnification.
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
35
Fig. 10. (a) Load–axial displacement curves of the triplex SiC composite tube showing a load drop phenomenon and (b) Weibull distribution of hoop strength at initial load drop.
SiC triplex as a LWR fuel cladding tube. Extensive microcracks throughout the SiCf/SiC composite layer could be developed right after a failure SiC inner layer which can lead to the release of fission products. Furthermore, it has been known that the irradiation-induced swelling introduces the considerable tensile stress in the SiC inner layer when the SiC triplex cladding tube is under the LWR operating condition [9]. This effect overwhelms the compressive stress developed by the thermal gradient, resulting in the failure of the SiC inner layer at much lower internal pressure [32]. Fig. 11 shows the fracture morphology of the triplex tube after a significant load drop at the maximum hoop strength. The triplex tube retains its tubular geometry, as shown in Fig. 11(a) and (b).
Large expansions of the SiC triplex tube accompany the extensive cracking of the inner surface of the specimens in which a polyurethane plug was located. The SiC inner layer returns to the original shape after the removal of internal pressure, while the composite layer was permanently deformed, which could be caused by matrix microcracking. Due to the existence of PyC at the interface between the SiC inner and composite layers, a significant debonding was observed around the extensive cracking region in the SiC inner layer. After the bulging of the tubes, the subsequent cracking in composite and outer layers propagates to the axial direction only at the region where polyurethane plug was located. Since the stress concentrated on the composite layer around cracks of the SiC inner
Fig. 11. Fracture morphology of the SiC triplex tube after failure: (a) and (b) computed tomography (CT) images, and SEM images (c) beyond polyurethane plug and (d) of fracture surface.
36
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
layer, the axial crack in the composite layer propagates along the main crack the inner layer [33]. Beyond polyurethane plug, the crack propagates along the winding direction of the fiber bundles in which the crack resistance is minimum, as shown in Fig. 11(c). A SiC outer layer was fractured during the failure of the composite layer. It is believed that the load sharing was active between the composite and outer layers, as well as the hoop strength is minimum at the outside of the tube [34]. Fig. 11(d) shows the microstructure of composite layer after failure. The pull-out process of the fibers in the densely infiltrated matrix phase effectively enhances the toughness. The delamination occurs at the inter-ply region of the composite layer which is the weakest point for the composite layer with a two-dimensional fiber lay-up architecture. 4. Conclusions The SiC triplex tubes consisting of a SiC inner layer, a SiCf/SiC composite layer, and a SiC outer layer were fabricated via the chemical vapor processes for the LWR fuel cladding application. The fiber volume fraction varied by adjusting the band width of fibers and the winding patterns. The hoop strength of the SiC triplex tubes tended to increase depending on the aligned fiber volume fraction of the composite layer and the winding angle of the fiber bundles from 45° to 65°. The Tyranno SA3-0.8k reinforced SiC triplex tube exhibited the highest hoop strength because the highest fiber volume fraction was achieved by a filament winding method. On the other hand, the fiber volume fraction of Hi Nicalon Type S was less than Tyranno SA3 because a diameter of Hi-Nicalon Type S was higher than Tyranno SA3 because the Hi-Nicalon Type S fiber is less flexible. The SiC triplex tubes had a quite high average ultimate hoop strength, 282.4 MPa with a relatively high Weibull modulus of 11.05. However a SiC inner layer was under maximum tensile stress and failed at the lower stress with the wide distribution. Extensive microcracks throughout the SiCf/SiC composite layer were developed after a failure SiC inner layer which may introduce uncertainties of the SiC triplex cladding. Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2012M2A8A5009818). References [1] G. Makham, R. Hall, H. Feinroth, Ceram. Eng. Sci. Proc. 33 (2013) 101–120. [2] K. Barrett, S. Bragg-Sitton, D. Galicki, Advanced LWR Nuclear Fuel Cladding System Development Trade-off Study, INL/EXT-12-27090, 2012.
[3] T. Cheng, J.R. Keiser, M.P. Brady, K.A. Terrani, B.A. Pint, J. Nucl. Mater. 427 (2012) 396–400. [4] K. Yueh, D. Carpenter, H. Feinroth, Nucl. Eng. Int. 55 (2010) 14–16. [5] Y. Katoh, T. Nozawa, L.L. Snead, K. Ozawa, H. Tanigawa, J. Nucl. Mater. 417 (2011) 400–505. [6] L.L. Snead, T. Nozawa, Y. Katoh, T.-S. Byun, S. Kondo, D.A. Petti, J. Nucl. Mater. 371 (2007) 329–377. [7] G. Youinou, R.S. Sen, Enhanced Accident Tolerant Fuels for LWRs – A Preliminary System Analysis, INL/EXT-13-30211, 2013. [8] H. Feinroth, E-J. Adv. Maint. 5 (2013) 128–131. [9] Y. Katoh, K.A. Terrani, L.L. Snead, Systematic Technology Evaluation Program for SiC/SiC Composite-based Accident-Tolerant LWR Fuel Cladding and Core Structures, ORNL/TM-2014/210, 2014. [10] W.-J. Kim, H.S. Hwang, J.Y. Park, J. Mater. Sci. Lett. 21 (2002) 733–735. [11] W.-J. Kim, H.S. Hwang, J.Y. Park, W.-S. Ryu, J. Mater. Sci. Lett. 22 (2003) 581–584. [12] J.-Y. Park, I.-H. Kim, Y.-I. Jung, H.-G. Kim, D.-J. Park, W.-J. Kim, J. Nucl. Mater. 433 (2013) 603–607. [13] Y. Katoh, L.L. Snead, T. Cheng, C. Shih, W.D. Lewis, T. Koyanagi, T. Hinoki, C.H. Henager Jr., M. Ferraris, J. Nucl. Mater. 448 (2014) 497–511. [14] T. Hinoki, Y. Katoh, A. Kohyama, Mater. Trans. 43 (2002) 617–621. [15] H. Kishimoto, Y. Katoh, A. Kohyama, J. Nucl. Mater. 307–311 (2002) 1130–1134. [16] Y. Katoh, K. Ozawa, T. Hinoki, Y. Choi, L.L. Snead, A. Hasegawa, J. Nucl. Mater. 417 (2011) 416–420. [17] Y. Katoh, T. Nozawa, L.L. Snead, K. Ozawa, H. Tanigawa, J. Nucl. Mater. 416 (2011) 400–405. [18] T. Koyanagi, K. Ozawa, T. Hinoki, K. Shimoda, Y. Katoh, J. Nucl. Mater. 448 (2014) 478–486. [19] Y. Katoh, M. Kotani, H. Kishimoto, W. Yang, A. Kohyama, J. Nucl. Mater. 289 (2001) 42–47. [20] S. Bragg-Sitton, Advanced LWR Nuclear Fuel Cladding System Development Technical Program Plan, INL/MIS-12-25696, 2012. [21] D. Shaw-Stewart, Mater. Des. 6 (1985) 140–144. [22] T.S. Byun, L. Lara-Curizio, R.A. Lowden, L.L. Snead, Y. Katoh, J. Nucl. Mater. 367–370 (2007) 653–658. [23] J.Y. Park, S.M. Kang, L.H. Park, W.-J. Kim, W.S. Ryu, Ceram. Eng. Sci. Proc. 24 (2003) 613–620. [24] W. Yang, H. Araki, A. Kohyama, C. Busabok, Q. Hu, H. Suzuki, T. Noda, Mater. Trans. 44 (2003) 1797–1801. [25] G. Newsome, L.L. Snead, T. Hinoki, Y. Katoh, D. Peters, J. Nucl. Mater. 371 (2007) 76–89. [26] L.J. Broutman, R.H. Krock, Composite Materials – Structural Design and Analysis: Part I, Academic Press, New York, 1975. [27] B.D. Agarwal, L.J. Broutman, Analysis and Performance of Fiber Composites, John Wiley & Sons, New York, 1990. [28] E.V. Morozov, Compos. Struct. 76 (2006) 123–129. [29] D. Kim, J. Lee, J.Y. Park, W.-J. Kim, J. Kor. Ceram. Soc. 50 (2013) 359–363 (in Korean). [30] P.D. Soden, R. Kitching, P.C. Tse, Y. Tsavalas, Compos. Sci. Technol. 46 (1993) 363–378. [31] M. Xia, K. Kemmochi, H. Takayanagi, Compos. Struct. 51 (2001) 273–283. [32] Y. Lee, M.S. Kazimi, A structural model for multi-layered ceramic cylinders and its application to silicon carbide cladding of light water reactor fuel, J. Nucl. Mater., 2014, accepted for publication. [33] J.D. Stempien, D.M. Carpenter, G. Kohse, M.S. Kazimi, Nucl. Technol. 183 (2013) 13–29. [34] H. Feinroth, M. Ales, E. Barringer, G. Kohse, D. Carpenter, R. Jaramillo, Mechanical Strength of CTP triplex SiC fuel clad tubes after irradiation in MIT research reactor under PWR coolant conditions, in: Proc. 33rd ISACC, 2013.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Fabrication and measurement of hoop strength of SiC triplex tube for nuclear fuel cladding applications Daejong Kim ⇑, Hyun-Geun Lee, Ji Yeon Park, Weon-Ju Kim Nuclear Materials Division, Korea Atomic Energy Research Institute, 989-111 Daedeok-daero, Yuseong-gu, Daejeon 305-353, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 June 2014 Accepted 26 November 2014 Available online 3 December 2014
a b s t r a c t The SiC ceramics are under investigation for the fuel cladding in the light water nuclear reactors because of its excellent high temperature strength and corrosion resistance against hot steam under the severe accident conditions. In this study, the SiC triplex tubes consisting of a SiC inner layer, a SiC/PyC/SiC intermediate layer, and a SiC outer layer were fabricated by the chemical vapor processes. The hoop strength and fracture behaviors of the SiC triplex tube were investigated. The SiC triplex tubes fabricated at the high ratio of H2/MTS had a quite high average strength with a relatively small standard deviation. The hoop strength of the composite tubes tends to increase with the volume fraction of the reinforced fibers. The highest fiber volume fraction was obtained using Tyranno SA3-0.8k with the dense winding patterns such as bamboo-like mosaic pattern, which resulted in the high hoop strength compared to other fibers of Tyranno SA3-1.6k and Hi-Nicalon Type S. Hoop strength also increased slightly as the winding angle increased from 45° to 65°. Fracture behaviors of the SiC triplex tube were investigated via the observation of microstructure of the failed samples. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction The current light water reactors (LWRs) have used Zr alloys as a nuclear fuel cladding due to its superior properties such as a low neutron absorption cross-section and a good neutron irradiation resistance under the operation conditions of the reactor. However, Zr alloys react rapidly with hot steam produced during severe accidents such as loss-of-coolant-accidents (LOCA), undergoing significant recession and degradation of mechanical strength, and producing massive hydrogen [1]. Therefore, there have been extensive efforts to develop accident tolerance fuel cladding materials to replace Zr alloys. There are several new design concepts of LWR fuel cladding, categorized by Zr-based concepts such as SiC or MAX phase coated Zr alloy, SiC composite + Zr alloy or Zr–Mo–Zr sandwich hybrid, and fully replacement concepts such as Fe–Cr–Al alloys, Mo alloys, SiC triplex, and stainless steels [2]. Most of the new concepts have considered a use of SiC ceramics as their constituent because the hydrogen liberation rate of SiC is hundreds times less than Zr alloys [1,3]. Furthermore, SiC and its composites have excellent high temperature mechanical properties, a low neutron absorption cross section, a high melting point, a good neutron irradiation resistance, and little irradiation creep, compared to Zr alloys [4–6]. ⇑ Corresponding author. Tel.: +82 42 868 4559; fax: +82 42 868 8549. E-mail address: [email protected] (D. Kim). http://dx.doi.org/10.1016/j.jnucmat.2014.11.117 0022-3115/Ó 2014 Elsevier B.V. All rights reserved.
Among various nuclear fuel concepts, the SiC triplex proposed by CTP Ltd. has been considered as one of the leading concepts because of its larger safety margins under beyond-design basis severe accident and high burnup capability [7,8]. However, there are still many technical issues such as understanding of statistical failure of the SiC triplex tubes, stress concentration on an inner SiC layer and failure due to swelling-induced stress, fission products release due to microcracking in a matrix SiC phase, chemical compatibility between coolant and cladding under operating conditions, development of the hermetic joining method of end-plug, and manufacturability of long cladding tubes and manufacturing defects [9–13]. The SiC triplex fuel cladding consists of a monolith SiC inner layer, a SiC/SiC composite intermediate layer, and a monolith SiC outer layer, as shown in Fig. 1. A primary function of the SiC inner layer is to prevent radioactive fission products from escaping the fuel into the coolant, the SiC/SiC composite layer enhances the strength of the cladding tube and provides better fracture toughness, and the SiC outer layer protects a composite layer against corrosive coolant water at high temperatures. In order to exhibit the best performance during neutron irradiation, each layer should be pure, well-crystallized and near-stoichiometric [14,15]. With regard to fiber selection only Tyranno SA3 and Hi-Nicalon Type S SiC fibers meet the properties required among various commercial products. In case of monolith SiC and SiC matrix phases, high quality SiC phases is usually obtained by
30
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
the chemical vapor processes. It has been recently reported that the chemical vapor infiltration (CVI)- and the nano-infiltration and transient eutectic phase (NITE)-processed SiC composites have best neutron irradiation resistance [16–18]. The composite produced by other process methods such as polymer impregnation and pyrolysis (PIP) and liquid silicon infiltration (LSI) contains impurities which would result in the poor performance during irradiation such as local swelling and strength degradation [19,20]. Therefore, in this study, the SiC triplex tubes which consisted of a monolith SiC inner layer, a SiC/PyC/SiC composite intermediate layer and a monolith SiC outer layer were fabricated by the chemical vapor processes such as CVD and CVI. Influences of filament winding methods, type of SiC fibers on hoop strength of the triplex tubes were investigated. The damage process during the hoop tests were examined via microstructure observation. 2. Experimental procedure 2.1. Fabrication of SiC triplex composite tube A monolith SiC inner layer was uniformly deposited with about 320.3 ± 6.4 lm by a chemical vapor deposition method using methyltricholorosilane (MTS, CH3SiCl3) onto the high purity graphite rods with 8.5 mm in diameter and a length of 100 mm. Then the SiC fibers were applied to a SiC-coated cylindrical mandrel by a filament winding method [21]. SiC fiber bundles (tows) consisting of 0.8–1.6k fibers were provided on a spool under a little tension which was generated only by the friction of the fiber supplying system. 1–2 helical plies of SiC fiber bundles covered the mandrel. The properties of SiC fibers are listed in Table 1. The commercial generation III SiC fibers such as Tyranno SA3 and Hi-Nicalon Type S were used in this study. They contain a small amount of impurities and they are crystallized and near-stoichiometric, so called nuclear grade SiC fibers. A fiber volume fraction was controlled by changing the band width of each fiber. For helical winding it is usual to lay fibers side by side, to give complete coverage as the winding progresses. Because of the difference in band width, the reciprocating motions of a transverse carriage for the complete coverage were varied from 12 to 21 times, which results in the change in fiber volume fraction and corresponding thickness of the composite intermediate layer in the range of 240–430 lm. As the input value of the band width is smaller than the actual value, accordingly the pattern is open which results in the less fiber volume fraction. The winding pattern was controlled by adjusting the position of second fiber trajectory. Fig. 2 shows four phase winding patterns wound with the winding angle of ±55° using Tyranno SA3-0.8k and schematic illustrations. Winding angles were varied to ±45°, ±55°, and ±65° using Tyranno SA3-0.8k with the 1-return winding pattern. After the filament winding, PyC was deposited as an interphase of the SiCf/SiC composite onto SiC fibers by dehydrogenation of CH4
at 1100 °C using by a CVD method. For the infiltration of SiC matrix phase, the isothermal CVI was processed at 1000 °C at the low pressure of 3.3 kPa to reduce density gradient to the radial direction of the tubular specimens. Finally a SiC outer layer was deposited with the same conditions to the matrix infiltration in sequent. The SiC outer layer exhibited the wide thick variation of 39.8 ± 18.6 lm. 2.2. Measurement of hoop strength The tubular specimens for hoop strength measurements were prepared by following procedure. The substrate graphite rod was burned out at 1200 °C for 30 min and 900 °C for 21 h in the atmospheric environment. Graphite ash remained on the inner surface of the composite tube was removed by air blowing and ultrasonic cleaning. The tubular specimens have a dimension of a length of 10 and 30 mm, an inner diameter of 8.5 mm, and an outer diameter of 9.6–10.1 mm. Then the polyurethane plug with 8.45 mm in diameter and a length of 8 and 22 mm was inserted into the tubular specimen with a length of 10 mm and 30 mm, respectively. Hoop strength was measured via internal pressurization at room temperature using screw driven universal testing machine (Instron 4465, load cell capacity = 5 kN). Axial pressure was applied with loading rate of 0.01 mm/s to the cylindrical polyurethane plug. A series of hoop tests were carried out at room temperature and each test was repeated 2–7 times. Weibull modulus of the hoop strength was obtained using 20 samples with 685 ± 41 lm thickness and a length of 10 mm. Radial displacement measurements were made by attaching four displacement transducers (Kyowa, Co. Ltd) to the central section of the outer tube surface. Hoop strength of tubular SiC composite specimens, rh was calculated by [22]
rh ¼
r2i P r 2o 1 þ r 2i r 2o r 2i
ð1Þ
where P is the internal pressure, ri and ro are the inner and outer radii of the tubular specimens, respectively. Contact strength was typically 50–100 N which was determined at the increasing point of radial displacement. 3. Results and discussion 3.1. Microstructure of SiC triplex tube Fig. 3 shows microstructures of a SiC triplex tube and its constituent layers of a SiC inner layer and a Tyranno SA3-reinforced SiCf/ SiC composite layer. The CVD b-SiC inner layer has columnar structure elongated to the growth direction, as shown in Fig. 3(b). The layer is dense without voids. The SiCf/SiC composite layer consists of SiC fibers, PyC interphase, and SiC matrix, as shown in Fig. 3(c). The PyC interphase of about 200 nm in thickness was deposited onto the SiC fibers. The intrabundle space within the fiber woven composite layer was densely filled with SiC by a CVI process. Some circumferential voids were formed in the SiC matrix but the fraction of voids is a quite low compared to typical CVI-processed SiCf/SiC composite [23]. The columnar grains of the SiC matrix are grown to the radial direction of the fibers. The grain size is much smaller than the SiC inner layer because SiC matrix was deposited at the lower temperature of 1000 °C. 3.2. Hoop strength of SiC triplex tube
Fig. 1. SiC triplex tube concept for LWR fuel cladding application.
3.2.1. Hoop strength and Weibull modulus Fig. 4 shows the Weibull statistical plot of ultimate hoop strength of SiC triplex tubes whose intermediate SiCf/SiC composite layer were infiltrated at 1000 °C at H2/MTS ratio of 20. The
31
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36 Table 1 Properties of SiC fibers used in this study. SiC fiber
Compositions (wt%)
Filament dia. (mm)
Filaments/yarn
Tensile strength (GPa)
Tensile modulus (GPa)
Ply no.
Tyranno SA3-0.8k Tyranno SA3-1.6k Hi-Nicalon Type S
68Si + 32C + 0.6Al 68Si + 32C + 0.6Al 69Si + 31C + 0.2O
7.5 7.5 12
800 1600 500
2.1 2.1 2.6
395 395 420
2 1 1
Fig. 2. (a) Four phase winding patterns and (b) schematic illustrations wound by a helical filament winding method using Tyranno SA3-0.8k with the winding angle of ±55°.
Fig. 3. (a) SEM microstructures of the SiC triplex tube and back scattered electron images showing microstructures of (a) a monolith SiC inner layer and (b) a Tyranno SA3 reinforced SiCf/SiC composite layer.
ultimate hoop strength for 20 samples was within the 235.0– 337.8 MPa. The SiC triplex tubes had a quite high average strength, 282.4 MPa, with a relatively small standard deviation, 44.3 MPa. It
has been known that CVD SiC has Weibull modulus of 5–11 depending on the volume of specimens and test methods. 2D plane-woven Tyranno SA or Hi Nicalon SiC fiber reinforced SiCf/
32
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
Fig. 4. Weibull statistical plot of ultimate hoop strength of SiC triplex tubes.
SiC composite has Weibull modulus of about 10 [17,6,24,25]. In case of filament wound composite tubes, the matrix density of the SiCf/SiC composite layer was higher than that of the 2D plane-woven composite which was measured to be about 2.92 g/ cm3. Therefore, Weibull modulus of the SiC triplex tubes was high to be 11.05. This value is comparable or higher than CVD SiC and plane-woven SiCf/SiC composite. Since all samples for Weibull statistics were taken from the same batch process, however, the deviation of strength can increase due to the difference in matrix density between batch processes. When considering that an orthotropic lamina composite with its principal material (L and T) axes oriented at an angle, h with the reference coordinate axes (x and y), stresses can be transformed from one set of axes to another [26,27]:
8 9 > < rL > =
9 38 > < rx > = 7 rT ¼ 6 4 sin2 h cos2 h 2 sinh cosh 5 ry > > > : > sLT ; sinh cosh sinhcos h cos2 h sin2 h : sxy ; 2
cos2 h
2
sin h
2sinhcos h
ð2Þ
If a stress, rx is applied to an orthotropic lamina, the longitudinal, shear, and transverse stress of the composite along the L and T directions can be calculated by the stress-transformation law:
rL ¼ rx cos2 h sLT ¼ rx sin h cos h rT ¼ rx sin2 h
ð3Þ ð4Þ ð5Þ
Based on the maximum stress theory associated with a weakest link phenomenon, the failure mechanism changes from tensile fraction of the fibers to matrix/interphase shear to matrix/interphase cleavage as the fiber angle increases from 0° to 90°. When a unidirectional composite which is loaded parallel to the fiber direction, with angled-ply orientation, and perpendicular to the fiber direction, the fiber volume fraction, Vf ideally has an effect on the normal and shearing stress as follows [27]:
rL ¼ rf V f þ rm ð1 V f Þ s s sLT ¼ GLT f V f þ m ð1 V f Þ Gf
rT ¼ rm
Gm
ET ð1 V 1=3 Þ Em
Assuming the uniaxial tensile stress is applied to the circumferential direction of the SiC triplex tube during the expanding plug test, the hoop strength of the SiC triplex tube with the fixed winding angle increased with the aligned fiber volume fraction, as shown in Fig. 5. All samples have almost same thickness of SiC inner with about 320 lm and outer layers with 40 lm. An increase in hoop strength could be contributed by a SiCf/SiC composite layer whose thickness increased from 250 to 420 lm with the fiber volume fraction. Therefore, it is important for obtaining higher hoop strength to increase the fiber volume fraction per unit composite volume. The highest fiber volume fraction was obtained when the composite layer was wound by Tyranno SA3-0.8k fiber which resulted in the highest hoop strength. In the case of the Hi-Nicalon Type S-reinforced triplex tubes, the fiber volume fraction is less than Tyranno SA3 that has the same winding architecture because a diameter of Hi-Nicalon Type S is higher than Tyranno SA3 and thus Hi-Nicalon Type S is less flexible. Although Hi-Nicalon Type S has higher tensile strength and tensile modulus (Table 1), its composite tube has similar hoop strength with the Tyranno SA3-reinforced triplex tube. 3.2.2. Effect of filament winding pattern and angle The winding pattern itself leads to the different strength and strain field [28]. Moreover, it has a significant effect on fiber volume fraction and void distribution between fiber bundles [29]. As shown in Fig. 2, the fiber bundle for the bamboo-like mosaic pattern such as 1-return and 2-return lays side by side and there is little interference between fiber bundles with ±h angles. Therefore, the void fraction of a SiCf/SiC composite layer after the CVI process was relatively low which was 10.3% for 1-return and 11.1% for 2-return. A small size of voids is distributed mainly at the intrabundle site of the fiber. Relatively large voids exist in the interbundle and the inter-ply region, as shown in Fig. 6(a) and (b). For the triplex tubes with 6-return and 9-return winding patterns, on the other hand, two or three times space of the band width exists between fiber bundles with +h angle. Fiber bundles with h angle are slightly lifted due to interference of +h angle fiber bundle, resulting in the formation of large interbundle voids, which results in an increase in thickness of a composite layer. As shown in Fig. 6(c) and (d), the fiber volume fraction is relatively low and large interbundle and inter-ply voids are formed after the CVI process. The void fractions of the composite layer were relatively high which were 16.8% and 14.9% for the composite layers with 6-return and 9-return, respectively. These large circumferential voids may be detrimental on the integrity of fuel cladding. When a SiC inner layer is failed, fission products can easily release through the microcracks in matrix phase and voids into the
ð6Þ ð7Þ ð8Þ
where rf and rm are the ultimate strength of the fiber and the matrix stress at the fiber fracture strain, respectively. E and G are elastic and shear modulus, respectively. As the fiber volume fraction, Vf increases, therefore the fibers carry more of the applied load at a low angle to the fiber direction and the strength of the composite increases with the aligned fiber volume fraction.
Fig. 5. Hoop strength of the SiC triplex tubes as a function of the fiber volume fraction.
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
33
Fig. 6. SEM microstructure of the SiC triplex tubes reinforced by Tyranno SA3-0.8k with the winding patterns of (a) 1-return, (b) 2-return, (c) 6-return, and (b) 9-return.
coolant. Furthermore, thermal conductivity of the SiC triplex tube will be decreased. Fig. 7 shows the hoop strength of the SiC triplex tubes with various winding patterns and winding angles which were reinforced by Tyranno SA3-0.8k. The hoop strength tends to increase with increasing the fiber volume fraction depending on the winding patterns, as shown in Fig. 7(a). However, the hoop strength of the tubes with 2, 6, 9-return winding patterns is higher than that of 1-return although the tube with 1-return winding pattern had highest fiber volume fraction. It is believed due to an increase in the fraction of the composite layer in the whole triplex tube. Fig. 7(b) shows the hoop strength of the triplex tubes reinforced by Tyranno SA3-0.8k with the different winding angle. Since fibers have superior mechanical properties along their lengths, higher orientations of the fibers provide the higher hoop strength of the composite tube [30,31]. Therefore, the hoop strength of the tube increases as the winding angle gets higher. However, the effect of winding angle on the hoop strength is less pronounced on the corresponding properties of 45° and 65° wound tubes, compared to conventional SiCf/SiC composites which could be caused by that the fraction of the fiber reinforced composite layer is not much in the triplex tube. 3.3. Fracture behavior Fig. 8(a) shows the applied load-axial displacement curves of the SiC triplex tubes. Load gradually increases until the polyurethane plug contacts with the tubular specimen at the initial stage, and it steeply increased after the contact. Especially some load drops are observed before failure of the tube which might be caused by the formation of the large axial crack. The stress–strain curve displays an initial almost linear up to almost a half of the curve, as shown in Fig. 8(b). The onset of non-linearity could be associated with matrix cracking and manifests itself in degradation
Fig. 7. Hoop strength of the SiC triplex tubes depending on (a) winding pattern and (b) winding angle.
34
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
(a)
(b) Fig. 8. (a) Typical load–axial displacement and (b) stress–radial displacement curves of the SiC triplex tubes.
in tube stiffness after a significant load drop and an increase in radial displacement. Fig. 9 shows the microstructure of the triplex tube interrupted right after a first load drop during the hoop test. Failure of the SiC inner layer obviously takes place, as shown in Fig. 9(a). However, the crack from the inner layer does not propagated immediately into the composite layer. In the SiC triplex tube, PyC with about 200 nm in thickness exists between the SiC inner and composite layers, which was formed during the deposition of PyC interphase on SiC fibers. When the crack propagates from inner layer to composite layer, therefore the PyC interphase become the preferred site for crack propagation. It indicates that the existence of a thin PyC layer play a role of an obstacle of crack propagation. However, after a failure of the SiC inner layer, the stress concentrated on a SiCf/SiC composite layer, followed by a large expansion of the quasi-ductile composite layer. Therefore, a number of intrabundle microcracks were developed primarily along the matrix/ fiber interface to the circumferential direction by shear stress at the initial fracture stage. A small population of the radial cracks was also observed, as shown in Fig. 9(c) and (d). Fig. 10 shows load-axial displacement curves of SiC triplex tubes with 10 mm length and Weibull distribution of the hoop strength at the initial load drop associated with a fracture of a SiC inner layer. Most of the triplex tube samples experienced load drop phenomena except for 3 samples. Hoop strength at the first load drop was in the wide range of 155.3–309.5 MPa. The average hoop strength was 226.95 MPa. Although these values may not represent the crack initiation, they imply the onset of a significant axial crack in the SiC inner layer. Weibull modulus for hoop strength at the first load drop was calculated to be 5.30 which is much lower than ultimate hoop strength. Failure of a SiC inner layer at the lower stress and the wide distribution of failure strength may have a detrimental effect on the reliability of the
Fig. 9. Microstructure of the SiC triplex tube after a first load drop: (a) a crack in a SiC inner layer and (b) a matrix crack in a SiCf/SiC composite layer at low magnification and (c) high magnification.
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
35
Fig. 10. (a) Load–axial displacement curves of the triplex SiC composite tube showing a load drop phenomenon and (b) Weibull distribution of hoop strength at initial load drop.
SiC triplex as a LWR fuel cladding tube. Extensive microcracks throughout the SiCf/SiC composite layer could be developed right after a failure SiC inner layer which can lead to the release of fission products. Furthermore, it has been known that the irradiation-induced swelling introduces the considerable tensile stress in the SiC inner layer when the SiC triplex cladding tube is under the LWR operating condition [9]. This effect overwhelms the compressive stress developed by the thermal gradient, resulting in the failure of the SiC inner layer at much lower internal pressure [32]. Fig. 11 shows the fracture morphology of the triplex tube after a significant load drop at the maximum hoop strength. The triplex tube retains its tubular geometry, as shown in Fig. 11(a) and (b).
Large expansions of the SiC triplex tube accompany the extensive cracking of the inner surface of the specimens in which a polyurethane plug was located. The SiC inner layer returns to the original shape after the removal of internal pressure, while the composite layer was permanently deformed, which could be caused by matrix microcracking. Due to the existence of PyC at the interface between the SiC inner and composite layers, a significant debonding was observed around the extensive cracking region in the SiC inner layer. After the bulging of the tubes, the subsequent cracking in composite and outer layers propagates to the axial direction only at the region where polyurethane plug was located. Since the stress concentrated on the composite layer around cracks of the SiC inner
Fig. 11. Fracture morphology of the SiC triplex tube after failure: (a) and (b) computed tomography (CT) images, and SEM images (c) beyond polyurethane plug and (d) of fracture surface.
36
D. Kim et al. / Journal of Nuclear Materials 458 (2015) 29–36
layer, the axial crack in the composite layer propagates along the main crack the inner layer [33]. Beyond polyurethane plug, the crack propagates along the winding direction of the fiber bundles in which the crack resistance is minimum, as shown in Fig. 11(c). A SiC outer layer was fractured during the failure of the composite layer. It is believed that the load sharing was active between the composite and outer layers, as well as the hoop strength is minimum at the outside of the tube [34]. Fig. 11(d) shows the microstructure of composite layer after failure. The pull-out process of the fibers in the densely infiltrated matrix phase effectively enhances the toughness. The delamination occurs at the inter-ply region of the composite layer which is the weakest point for the composite layer with a two-dimensional fiber lay-up architecture. 4. Conclusions The SiC triplex tubes consisting of a SiC inner layer, a SiCf/SiC composite layer, and a SiC outer layer were fabricated via the chemical vapor processes for the LWR fuel cladding application. The fiber volume fraction varied by adjusting the band width of fibers and the winding patterns. The hoop strength of the SiC triplex tubes tended to increase depending on the aligned fiber volume fraction of the composite layer and the winding angle of the fiber bundles from 45° to 65°. The Tyranno SA3-0.8k reinforced SiC triplex tube exhibited the highest hoop strength because the highest fiber volume fraction was achieved by a filament winding method. On the other hand, the fiber volume fraction of Hi Nicalon Type S was less than Tyranno SA3 because a diameter of Hi-Nicalon Type S was higher than Tyranno SA3 because the Hi-Nicalon Type S fiber is less flexible. The SiC triplex tubes had a quite high average ultimate hoop strength, 282.4 MPa with a relatively high Weibull modulus of 11.05. However a SiC inner layer was under maximum tensile stress and failed at the lower stress with the wide distribution. Extensive microcracks throughout the SiCf/SiC composite layer were developed after a failure SiC inner layer which may introduce uncertainties of the SiC triplex cladding. Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2012M2A8A5009818). References [1] G. Makham, R. Hall, H. Feinroth, Ceram. Eng. Sci. Proc. 33 (2013) 101–120. [2] K. Barrett, S. Bragg-Sitton, D. Galicki, Advanced LWR Nuclear Fuel Cladding System Development Trade-off Study, INL/EXT-12-27090, 2012.
[3] T. Cheng, J.R. Keiser, M.P. Brady, K.A. Terrani, B.A. Pint, J. Nucl. Mater. 427 (2012) 396–400. [4] K. Yueh, D. Carpenter, H. Feinroth, Nucl. Eng. Int. 55 (2010) 14–16. [5] Y. Katoh, T. Nozawa, L.L. Snead, K. Ozawa, H. Tanigawa, J. Nucl. Mater. 417 (2011) 400–505. [6] L.L. Snead, T. Nozawa, Y. Katoh, T.-S. Byun, S. Kondo, D.A. Petti, J. Nucl. Mater. 371 (2007) 329–377. [7] G. Youinou, R.S. Sen, Enhanced Accident Tolerant Fuels for LWRs – A Preliminary System Analysis, INL/EXT-13-30211, 2013. [8] H. Feinroth, E-J. Adv. Maint. 5 (2013) 128–131. [9] Y. Katoh, K.A. Terrani, L.L. Snead, Systematic Technology Evaluation Program for SiC/SiC Composite-based Accident-Tolerant LWR Fuel Cladding and Core Structures, ORNL/TM-2014/210, 2014. [10] W.-J. Kim, H.S. Hwang, J.Y. Park, J. Mater. Sci. Lett. 21 (2002) 733–735. [11] W.-J. Kim, H.S. Hwang, J.Y. Park, W.-S. Ryu, J. Mater. Sci. Lett. 22 (2003) 581–584. [12] J.-Y. Park, I.-H. Kim, Y.-I. Jung, H.-G. Kim, D.-J. Park, W.-J. Kim, J. Nucl. Mater. 433 (2013) 603–607. [13] Y. Katoh, L.L. Snead, T. Cheng, C. Shih, W.D. Lewis, T. Koyanagi, T. Hinoki, C.H. Henager Jr., M. Ferraris, J. Nucl. Mater. 448 (2014) 497–511. [14] T. Hinoki, Y. Katoh, A. Kohyama, Mater. Trans. 43 (2002) 617–621. [15] H. Kishimoto, Y. Katoh, A. Kohyama, J. Nucl. Mater. 307–311 (2002) 1130–1134. [16] Y. Katoh, K. Ozawa, T. Hinoki, Y. Choi, L.L. Snead, A. Hasegawa, J. Nucl. Mater. 417 (2011) 416–420. [17] Y. Katoh, T. Nozawa, L.L. Snead, K. Ozawa, H. Tanigawa, J. Nucl. Mater. 416 (2011) 400–405. [18] T. Koyanagi, K. Ozawa, T. Hinoki, K. Shimoda, Y. Katoh, J. Nucl. Mater. 448 (2014) 478–486. [19] Y. Katoh, M. Kotani, H. Kishimoto, W. Yang, A. Kohyama, J. Nucl. Mater. 289 (2001) 42–47. [20] S. Bragg-Sitton, Advanced LWR Nuclear Fuel Cladding System Development Technical Program Plan, INL/MIS-12-25696, 2012. [21] D. Shaw-Stewart, Mater. Des. 6 (1985) 140–144. [22] T.S. Byun, L. Lara-Curizio, R.A. Lowden, L.L. Snead, Y. Katoh, J. Nucl. Mater. 367–370 (2007) 653–658. [23] J.Y. Park, S.M. Kang, L.H. Park, W.-J. Kim, W.S. Ryu, Ceram. Eng. Sci. Proc. 24 (2003) 613–620. [24] W. Yang, H. Araki, A. Kohyama, C. Busabok, Q. Hu, H. Suzuki, T. Noda, Mater. Trans. 44 (2003) 1797–1801. [25] G. Newsome, L.L. Snead, T. Hinoki, Y. Katoh, D. Peters, J. Nucl. Mater. 371 (2007) 76–89. [26] L.J. Broutman, R.H. Krock, Composite Materials – Structural Design and Analysis: Part I, Academic Press, New York, 1975. [27] B.D. Agarwal, L.J. Broutman, Analysis and Performance of Fiber Composites, John Wiley & Sons, New York, 1990. [28] E.V. Morozov, Compos. Struct. 76 (2006) 123–129. [29] D. Kim, J. Lee, J.Y. Park, W.-J. Kim, J. Kor. Ceram. Soc. 50 (2013) 359–363 (in Korean). [30] P.D. Soden, R. Kitching, P.C. Tse, Y. Tsavalas, Compos. Sci. Technol. 46 (1993) 363–378. [31] M. Xia, K. Kemmochi, H. Takayanagi, Compos. Struct. 51 (2001) 273–283. [32] Y. Lee, M.S. Kazimi, A structural model for multi-layered ceramic cylinders and its application to silicon carbide cladding of light water reactor fuel, J. Nucl. Mater., 2014, accepted for publication. [33] J.D. Stempien, D.M. Carpenter, G. Kohse, M.S. Kazimi, Nucl. Technol. 183 (2013) 13–29. [34] H. Feinroth, M. Ales, E. Barringer, G. Kohse, D. Carpenter, R. Jaramillo, Mechanical Strength of CTP triplex SiC fuel clad tubes after irradiation in MIT research reactor under PWR coolant conditions, in: Proc. 33rd ISACC, 2013.