- Email: [email protected]
Effect of reinforcement geometry on the mechanical properties of SiCwAl2O3 composites
Composites Engineering. Vol. 2, No. 4, pp. 239-247, 1992.
0961-9526/92 SS.OO+ .tXl 0 1992 Pergamon Press Ltd
Printed in Great Britain.
EFFECT OF REINFORCEMENT GEOMETRY ON THE MECHANICAL PROPERTIES OF SiC,/A1203 COMPOSITES SIRAJUS SALEKIN, ANWARUL HAQUE, NURUL HUQ, J. S. COPES,’ HASSAN MAHFUZ and SHAIK JEELANI Materials Research Laboratory, Tuskegee University, Tuskegee, AL 36088, U.S.A. (Received 1 August 1991; final version accepted 2 January 1992) Abstract-The mechanical properties of ceramic composites change due to variations in the size and geometry of the reinforcement. In the present work the effects of Sic whisker (SIC,) and Sic platelets (Sic,) on the mechanical properties of SiC/Al,O, composites have been investigated. The flexural properties of both whisker- and platelet-reinforced SiC/Al,O, composites were measured using the four-point bend test. Fracture toughness was determined for both straight- and chevronnotched specimens. A scanning electron microscopy examination of the fractured surfaces was conducted to perform failure analyses and understand toughening mechanisms. The results of the investigation indicate that SiC,/Al,O, composites provides higher flexural strength and fracture toughness compared to SiC,/Al,Os composites. During hot pressing, the control of the orientation of the Sic, is extremely difficult and it appears to form a porous matrix structure which substantially lowers the strength. Moreover, it has been observed that the low I/d ratio of SIC, has a major influence on the fracture toughness.
1. INTRODUCTION
The mechanical properties of monolithic ceramics can be improved significantly by the composite approach. Although the composite properties primarily depend on the combination of materials, the shape and microstructural arrangement of the constituents have a significant impact on the mechanical properties of the composites. Whiskers, particles, platelets, fibers and fillers are most commonly used as reinforcements in structural ceramics. The use of whiskers and platelets in ceramic matrix composites has different implications in terms of orientation, aspect ratio and interface characteristics that control the structural behavior of the composites. Ceramics, being brittle materials, possess low fracture toughness, restricting their utilization in many structural applications. There has been increased interest in the strengthening and toughening of ceramics by the incorporation of particulates or whiskers within the microstructure (Jenkins et al., 1987; Wei and Becher, 1985; Homeny et al., 1987; Neegaard and Homeny, 1989; Claussen, 1976). It has been found that both the size and concentration of the dispersed particles influence the fracture toughness (Bulijan and Sarin, 1987; Bulijan et al., 1987). Larger dispersed particles contribute to greater toughness than the smaller sizes. The micromechanical residual stresses developed in SiC/Al,O, composites at high temperature appear to be highly dependent on the geometry of the reinforcing inclusion (Zhuang and Bradt, 1989). Crack deflection, fiber breakage, fiber debonding, fiber pullout and wake, etc., are various mechanisms by which whisker or platelets help in toughening the composites (Becher et al., 1988; Sing et al., 1989; Marshall and Ritter, 1987). It has also been observed that rod-shaped particles are more effective as toughening agents compared to disk- or spherical-shaped particles. Considering these various aspects of Sic platelets and Sic whiskers as reinforcements, the present study was undertaken to investigate the influence of both Sic whiskers and platelets in reinforcing monolithic A&O,. The current research investigated the improvement of mechanical properties, namely the flexural strength and fracture toughness, due to these reinforcements.
‘Senior Development Chemist, 3M Corporation,
St Paul, MN 551441000, U.S.A. 239
240
S. SfiEmN et al. 2. EXPERIMENTAL
WORK
2.1. Specimen preparation The raw materials used in the preparation of billets were A&O3 matrix, Sic whiskers and SIC platelets. The A1,09 was in the form of powder (EBON A) and was supplied by Cercom.+ The Sic whiskers and Sic platelets were manufactured by Tokai Carbon* and American Matrix Inc.$, respectively. The Sic whiskers or platelets and A1203 powder were blended by liquid processing to form a slurry. This was then dried under agitation to produce a homogeneous powder. The total reinforcement content was maintained at 30% by volume. Hot pressing of the SiC/A120, mixture was carried out at Cercom and at Industrial Ceramic Technology,” in an inert atmosphere of nitrogen using graphite dies. The maximum hot pressing temperature was 1775°C and the maximum pressure applied was 4500 psi. For the fabrication of monolithic alumina billets, the hot pressing was carried out at a maximum temperature of 1475°C and the maximum pressure applied was 2OOOpsi. Billets of both monolithic A&O, and SiC/Al,Os composites were prepared for comparing the flexural strength and fracture toughness values. All the specimens for flexural strength and fracture toughness were machined and cut as specified in Figs l-3. It is well known that whiskers tend to align in the direction perpendicular to the hot-pressing axis. Accordingly, attention was given whilst cutting all the specimens to make certain that the loading axis remained perpendicular to the whisker length. In the case of platelets such a controlled orientation was, however, extremely difficult to achieve through hot pressing.
Fig. 1. Flexural test specimen.
l-----~mW Fig. 2. Straight-notched specimen.
PR
PR
I
g=yq-ijl Fig. 3. Chevron-notched
specimen.
’ Cercom Incorporated, 1960 Watson Way, Vista, CA 92083, U.S.A. ‘Tokai Carbon, American Inc., 375 Park Avenue, Suite 3802, NY 10152, U.S.A. ‘American Matrix Inc., Box 23556, Knoxville, TN 37933, U.S.A. ‘Industrial Ceramic Technology, 37 Enterprise Drive, Ann Arbor, MI 48103, U.S.A.
Effect of reinforcement geometry on composites
2.2. Measurement
241
of flexural strength and fracture toughness
The flexural strength of both whisker- and platelet-reinforced SiC/Al,03 composites and monolithic A&O, specimens were determined using the four-point bend test. Ten specimens of each type were tested to obtain a meaningful Weibull analysis. A ceramic bend fixture with 40 mm outer span and 20 mm inner span was used to conduct the tests. The tests were performed in an Instron load frame with a Series IX data acquistition system. The rate of cross-head speed was maintained at 0.02 in. min-‘. The flexural strength was calculated according to Whitney et al. (1982) and Zweben et al. (1979). The fracture toughness of both whisker- and platelet-reinforced SiC/AlzOs composites and monolithic A&O, was measured with two specimen configurations, straight-notched beam (SENB) and chevron-notched beam (CHNB). The chevron-notch configuration is shown in Fig. 4, and is characterized by the following parameters: (1) the notch length at the specimen surface, a, ; (2) the specimen width, W, and (3) the thickness, B. A total of 60 specimens were tested with approximately 10 in each category. In both specimen configurations, the tests were carried out in an Instron load frame with a crosshead speed of 0.02 in. min-‘. Due to the low-carrying capacity of both the notched specimens, a full-scale load of 50 lbs was selected for the test. Fracture toughnesses for the straight- and chevron-notched specimens were determined according to Munz et al. (1980a, b).
0
-------A Fig. 4. Plan view of the general chevron-notched fracture toughness sample with nomenclature. 3. RESULTS AND DISCUSSION
The flexural test is commonly employed in strength characterization of ceramic matrix composites due to the problem of grippping and aligning the specimen during the tensile test. The complexity in performing a high-temperature test with tensile fixture and the cost of tensile specimens are major concerns in characterizing ceramic matrix composites, but the common practice of performing a bend test appears to give a higher strength than the data obtained from a standard tensile coupon for both polymer and ceramic matrix composites (Whitney et al., 1982; Lamicq et al., 1986). Flaws existing in a specimen are subjected to a uniform maximum stress field during tensile loading. However, during flexure, the stress field varies almost linearly from the neutral axis of the beam cross-section. This variation in the stress field reduces the possibility of occurrence of flaws within the maximum stress field, and hence gives a higher strength. Again in bend tests, three-point loading produces a higher strength as compared to four-point loading. This is because the smaller volume under maximum stress gives a higher local strength. This difference in strength suggests the importance of finding the ratios for characteristic flexural strength to tensile strength using Weibull statistical strength theory for brittle materials (Weibull, 1939). Since Weibull statistics contain a volume term, they can predict the effect of volume on the strength parameter. The fracture toughness Ki, , measured from the critical stress-intensity factor K,, , is applied to linear elastic fracture mechanics. Ceramic materials behave in a linear elastic manner at least at low temperature, which reasonably justifies the measurement of the fracture toughness through the K1, value. The four-point bend method with both straightand chevron-notched specimens appears to be a promising candidate for standard K,,
242
S. SALEIUN
Flexurul
et al.
strength
( MPa 1
Fig. 5. Weibull plot of flexural strength data.
measurement techniques. Despite such a presumption, the results show a discrepancy in the measured values due to the influence of the specimen dimensions, loading rates and loading fixtures, showing the need for a standard testing technique in order to obtain a true K,, value for ceramic materials. Data for the flexural strength of A&O,, SiC,/Al,O, and SiC,/A1203 were analyzed using the Weibull distribution curves shown in Fig. 5, to establish the probability of failure occurrence. It is evident from Fig. 5 that the flexural strength of single-phase monolithic A1203 has been increased almost one and half times due to the incorporation of Sic whiskers, whereas in the case of Sic platelets the strength of the composite has been decreased. Most of the flexural strength data in Fig. 5 show linearity and a well-distributed fit in the Weibull plot, except for few in the case of A120,. The bad fit represents the material behavior and the quality of the data. The data marked with an asterisk for A120, show a material defect due to a significantly lower flexural strength. The value of the shape factor parameter of monolithic A&O3 would be higher if these data were omitted from the plot. A significantly higher value for the shape factor parameter denotes less scatter in the data of the flexural strength of SiC,/A1203. The difficulty with the hot-pressing technique as regards the orientation of the platelets is easily recognized from the SEM micrographs shown in Figs 6-8. To obtain the optimum strength properties in platelet-reinforced composites, flatness and freedom from surface and edge cracks are essential, but a high degree of parallelism in platelet distribution is more important than the uniformity of the particles themselves. The lack of such a parallelism in the platelet distribution is revealed by the SEM micrographs. Also, whenever a multiphase material is processed at elevated temperatures, the differences in the thermal expansion coefficients and elastic modulii of the phases result in residual stresses upon cooling to room temperature. These residual thermomechanical stresses have always been of interest from both the strengthening and the weakening perspective. The geometry of the inclusion, expressed as the aspect ratio (length/diameter or Nd) directly affects the magnitude of this internal stresses (Zhuang and Bradt, 1989). This ratio is a highly significant factor which can either amplify or reduce the thermal and elastic effects on the internal stresses. When it increases, the stress inside the inclusion parallel to the 1 axis significantly increases, and the tangential stress parallel to the I axis at the interface of the inclusion decreases. As a result, an inclusion with a whisper shape will carry more unidirectional load than one with a spherical shape. The results of the present study show that composites reinforced with platelets have a very low flexural strength and the low I/d ratio of the platelets is a significant factor in causing higher residual micromechanical stresses. However, an experimental investigation of such
Effect of reinforcement geometry on composites
Fig. 6. Scanning electron micrograph of fractured surface of SiC,/Al,O, .
243
Fig. 7. Scanning electron micrograph of fractured surface of SiC,/Al,O, (straight-notched specimen).
Fig. 8. Scanning electron micrograph of fractured surface of SiC,/Al,O, (chevron-notched specimen).
Effect of reinforcement geometry on composites
245
residual stresses is necessary before drawing any conclusions. The porosity due to incomplete densification is also of critical concern in platelet composites. These defects may be due to improper platelet/powder blending or because of harvesting of Sic platelets prior to mixing with Al,O, . Figures 9 and 10 show the Weibull distribution of the fracture toughness of monolithic A&O,, SiC,/A120, and SiC,/Al,O, composites obtained using the four-point bend test for straight- and chevron-notched specimens, respectively. A comparison of the Weibull characteristic values with the average value for both the flexural strength and facture toughness is presented in Table 1. It is observed that the K,, of A&O3 is increased with the incorporation of Sic whiskers and has a maximum value of 5.09 MPa 6 for straight-notched specimens and 7.69 for chevron-notched specimens. This observed Table 1. Comparison
of Weibull
characteristic lives and average vdues of flexural strength and fracture toughness
Flexural strength 4-point bend (MM
Materials
AW, Sic, /Al,O, Sic, /A&O,
Weibull characteristic life 391.46 620.72 122.83
Fracture toughness 4-point bend’ (MPa Ml”)
Average
Weibull characteristic life
370.22 608.59 119.85
4.27 5.17 3.53
Fracture toughness 4-point bend’ (MPa M”‘)
Average
Weibull characteristic life
Average
4.23 5.09 3.21
5.97 7.86 4.83
5.63 7.69 4.65
‘Straight notch. r Chevron notch.
increase is due to changes in the microstructure. It is observed that the average values are lower than the Weibull characteristic values. It is also evident that the chevron-notched specimens show higher K,, values than those of the streight-notched specimens for both Sic, and Sic, composites. The addition of whiskers to monolithic A&O3 results in crack deflection, whisker pullout and possibly crack bridging, as evident from the SEM micrographs (Fig. 6) of the fractured surface of the SiC,/A1203 specimen. Whisker pullout is detected in both the fractured surfaces. Holes are observed where whiskers dislodged prior to fracture, and in some cases whiskers are observed to be protruding above the fractured surface.
K,,
(MPaM’?
Fig. 9. Weibull plot of fracture toughness data (straight-notched COE 2:4-B
specimen).
246
2
J IO
Fig. 10. Weibull plot of fracture toughness data (chevron-notched specimen).
The state of stress at the matrix/whisker interface plays an important role in the toughening mechanisms. Due to the thermal coefficient expansion mismatch of A1203 and Sic,, large compressive stresses are developed in the Sic whiskers and tensile stresses are developed in the Al,O, matrix. Since the whiskers are in compression and the matrix is in tension at the interface, the main propagating crack is attracted to the tensile region and is deflected along the interface. It is clear from Figs 9 and 10 that the fracture toughness values of the SiC,/A1z03 composites are significantly lower than that of monolithic A&O,. The reasons given for the discrepancy in flexural strengths are also applicable to the fracture toughness. The SEM micrographs of platelet composites (Figs 7 and 8) show less evidence of any commonly implied toughening mechanisms observed in whisker composites. Analytical investigations have shown that the degree of toughening depends on the shape and therefore on the surface area of the dispersoid (Shalek et al., 1986). Both analytical and experimental investigations of the failure mechanisms with different sizes of platelets are necessary for a complete understanding of the effects of such reinforcements. 4. CONCLUSIONS
From the results of the investigation (1)
(2) (3)
the following conclusions can be drawn.
The flexural strength and fracture toughness of monolithic A1z03 increase under the incorporation of Sic whiskers. In the case of platelets both the strength and toughness of the composite appear to decrease, in comparison to monolithic AlzOs . It appears that chevron-notched specimens provide higher K,, values that those obtained using straight-notched specimens, for both Sic, and Sic, composites. Scanning electron microscopy reveals that there is lack of parallelism in platelet orientation which necessitates the development of a technique to control the platelet orientation in processing.
Acknowledgements-The authors would like to acknowledge the support from the Office of Naval Research through contract number NO0 14-86-k-0765 and the United States Air Force, Office of Scientific Research through contract number F-49620-20-89-C-0016DEF for this research. REFERENCES Becher, P. F., Hsueh, P. and Tiegs, T. N. (1988). Toughening behavior in whisker-reinforced ceramic matrix composite. J. Am. Ceram. Sot. 71(12), 1056-1061. Bulijan, S. T., Baldani, J. G. and Huckabee, M. L. (1987). Si,N,/SiC composites. Am. Ceram. Sot. Bull. t%(2), 347-352.
Effect of reinforcement geometry on composites
241
Bulijan, S. T. and Sarin, V. K. (1987). Silicon nitride-based composites. Composites M(2). 99-106. Claussen, N. (1976). Fracture toughness of A&O, with an unstabilized ZrO, dispersed phase. J. Am. Cer. Sot. 59(1-2), 49-51. Homeny, J., Vaughn, W. L. and Ferber, M. K. (1987). Processing and mechanical properties of SiCWhisker-A&O,-matrix composites. Am. Ceram. Sot. Bull. 66(2), 333-338. Jenkins, M. G., Kobayashi, A. S., White, K. W. and Bradt, R. C. (1987). Crack initiation and arrest in a Sic whisker/Al,O, matrix composite. J. Am. Ceram. Sot. 70(6), 363-395. Lamicq, P. J., Bernhart, G. A., Dauchier, M. M. and Mace, J. G. (1986). Sic/Sic composite ceramics. Am. Ceramic Sot. Bull. 65(2), 336-338. Marshall, D. B. and Ritter, J. E. (1987). Reliability of advanced structural ceramics and ceramic matrix composites-A review. Ceram. Bull. 66(2), 309-317. Munz, D., Busbey, R. T. and Shannon, J. L. (1980a). Fracture toughness determination of alumina using fourpoint bend specimens with straight-through and chevron notches. J. Am. Ceram. Sot. 63(5-6). Munz, D. J., Shannon, J. L. and Busbey, R. T. (1980b). Fracture toughness calculation from maximum load in four-point bend tests of chevron notch specimens. Int. J. Fracture 16R, 137-141. Neegaard, L. J. and Homeny, J. (1989). Mechanical properties of beta-silicon nitride whisker/silicon nitride matrix composites. Technical Report UIUC-NCCMR-89-0008, Department of Material Science and Engineering, University of Illinois, Urbana-Champaign. Shalek, P. D., Petrovic, J. J., Hurley, G. F. and Gac, F. D. (1986). Hot-pressed Sic whisker/Si,N, matrix composites. Am. Ceram. Sot. Bull 65(2), 351-356. Sing, J. P., Smith, S. and Scattergood, R. 0. (1989). Observation on the toughening of Also,-SiC composites. Materials and Components Technology Division, Argonne National Laboratory, Argonne, IL. Tiegs, T. N. and Becher, P. F. (1985). Whisker-reinforced ceramic composites. Am. Ceram. Sot. Bull. 64(2), 298-304. Wei, G. C. and Becher, P. F. (1985). Development of Sic whisker reinforced ceramics. Am. Ceram. Sot. Bull. 65(2), 298-304. Weibull, W. (1939). A statistical theory of the strength of materials. Ing. Vetenskaps Akad., Hand/., No. 151, pp. 5-45. Whitney, J. M., Daniel, I. M. and Pipes, B. (1982). Experimental Mechanics of Fiber Reinforced Composite Materials. Society for Experimental Stress Analysis, Monograph No. 4. Zhuang, L. and Bradt, R. C. (1989). Micromechanical stresses in Sic-reinforced A&O, composites. J. Am. Ceram. Sot. 72(l), 70-77. Zweben, C., Smith, W. S. and Wardle, M. W. (1979). Test methods for fiber tensile strength, composite flexure modulus, and properties of fabric-reinforced laminates. In Composite Materials: Testing and Deisign (Fifth Conf.), ASTM STP 674, pp. 228-262. ASTM, Philadelphia, PA.
0961-9526/92 SS.OO+ .tXl 0 1992 Pergamon Press Ltd
Printed in Great Britain.
EFFECT OF REINFORCEMENT GEOMETRY ON THE MECHANICAL PROPERTIES OF SiC,/A1203 COMPOSITES SIRAJUS SALEKIN, ANWARUL HAQUE, NURUL HUQ, J. S. COPES,’ HASSAN MAHFUZ and SHAIK JEELANI Materials Research Laboratory, Tuskegee University, Tuskegee, AL 36088, U.S.A. (Received 1 August 1991; final version accepted 2 January 1992) Abstract-The mechanical properties of ceramic composites change due to variations in the size and geometry of the reinforcement. In the present work the effects of Sic whisker (SIC,) and Sic platelets (Sic,) on the mechanical properties of SiC/Al,O, composites have been investigated. The flexural properties of both whisker- and platelet-reinforced SiC/Al,O, composites were measured using the four-point bend test. Fracture toughness was determined for both straight- and chevronnotched specimens. A scanning electron microscopy examination of the fractured surfaces was conducted to perform failure analyses and understand toughening mechanisms. The results of the investigation indicate that SiC,/Al,O, composites provides higher flexural strength and fracture toughness compared to SiC,/Al,Os composites. During hot pressing, the control of the orientation of the Sic, is extremely difficult and it appears to form a porous matrix structure which substantially lowers the strength. Moreover, it has been observed that the low I/d ratio of SIC, has a major influence on the fracture toughness.
1. INTRODUCTION
The mechanical properties of monolithic ceramics can be improved significantly by the composite approach. Although the composite properties primarily depend on the combination of materials, the shape and microstructural arrangement of the constituents have a significant impact on the mechanical properties of the composites. Whiskers, particles, platelets, fibers and fillers are most commonly used as reinforcements in structural ceramics. The use of whiskers and platelets in ceramic matrix composites has different implications in terms of orientation, aspect ratio and interface characteristics that control the structural behavior of the composites. Ceramics, being brittle materials, possess low fracture toughness, restricting their utilization in many structural applications. There has been increased interest in the strengthening and toughening of ceramics by the incorporation of particulates or whiskers within the microstructure (Jenkins et al., 1987; Wei and Becher, 1985; Homeny et al., 1987; Neegaard and Homeny, 1989; Claussen, 1976). It has been found that both the size and concentration of the dispersed particles influence the fracture toughness (Bulijan and Sarin, 1987; Bulijan et al., 1987). Larger dispersed particles contribute to greater toughness than the smaller sizes. The micromechanical residual stresses developed in SiC/Al,O, composites at high temperature appear to be highly dependent on the geometry of the reinforcing inclusion (Zhuang and Bradt, 1989). Crack deflection, fiber breakage, fiber debonding, fiber pullout and wake, etc., are various mechanisms by which whisker or platelets help in toughening the composites (Becher et al., 1988; Sing et al., 1989; Marshall and Ritter, 1987). It has also been observed that rod-shaped particles are more effective as toughening agents compared to disk- or spherical-shaped particles. Considering these various aspects of Sic platelets and Sic whiskers as reinforcements, the present study was undertaken to investigate the influence of both Sic whiskers and platelets in reinforcing monolithic A&O,. The current research investigated the improvement of mechanical properties, namely the flexural strength and fracture toughness, due to these reinforcements.
‘Senior Development Chemist, 3M Corporation,
St Paul, MN 551441000, U.S.A. 239
240
S. SfiEmN et al. 2. EXPERIMENTAL
WORK
2.1. Specimen preparation The raw materials used in the preparation of billets were A&O3 matrix, Sic whiskers and SIC platelets. The A1,09 was in the form of powder (EBON A) and was supplied by Cercom.+ The Sic whiskers and Sic platelets were manufactured by Tokai Carbon* and American Matrix Inc.$, respectively. The Sic whiskers or platelets and A1203 powder were blended by liquid processing to form a slurry. This was then dried under agitation to produce a homogeneous powder. The total reinforcement content was maintained at 30% by volume. Hot pressing of the SiC/A120, mixture was carried out at Cercom and at Industrial Ceramic Technology,” in an inert atmosphere of nitrogen using graphite dies. The maximum hot pressing temperature was 1775°C and the maximum pressure applied was 4500 psi. For the fabrication of monolithic alumina billets, the hot pressing was carried out at a maximum temperature of 1475°C and the maximum pressure applied was 2OOOpsi. Billets of both monolithic A&O, and SiC/Al,Os composites were prepared for comparing the flexural strength and fracture toughness values. All the specimens for flexural strength and fracture toughness were machined and cut as specified in Figs l-3. It is well known that whiskers tend to align in the direction perpendicular to the hot-pressing axis. Accordingly, attention was given whilst cutting all the specimens to make certain that the loading axis remained perpendicular to the whisker length. In the case of platelets such a controlled orientation was, however, extremely difficult to achieve through hot pressing.
Fig. 1. Flexural test specimen.
l-----~mW Fig. 2. Straight-notched specimen.
PR
PR
I
g=yq-ijl Fig. 3. Chevron-notched
specimen.
’ Cercom Incorporated, 1960 Watson Way, Vista, CA 92083, U.S.A. ‘Tokai Carbon, American Inc., 375 Park Avenue, Suite 3802, NY 10152, U.S.A. ‘American Matrix Inc., Box 23556, Knoxville, TN 37933, U.S.A. ‘Industrial Ceramic Technology, 37 Enterprise Drive, Ann Arbor, MI 48103, U.S.A.
Effect of reinforcement geometry on composites
2.2. Measurement
241
of flexural strength and fracture toughness
The flexural strength of both whisker- and platelet-reinforced SiC/Al,03 composites and monolithic A&O, specimens were determined using the four-point bend test. Ten specimens of each type were tested to obtain a meaningful Weibull analysis. A ceramic bend fixture with 40 mm outer span and 20 mm inner span was used to conduct the tests. The tests were performed in an Instron load frame with a Series IX data acquistition system. The rate of cross-head speed was maintained at 0.02 in. min-‘. The flexural strength was calculated according to Whitney et al. (1982) and Zweben et al. (1979). The fracture toughness of both whisker- and platelet-reinforced SiC/AlzOs composites and monolithic A&O, was measured with two specimen configurations, straight-notched beam (SENB) and chevron-notched beam (CHNB). The chevron-notch configuration is shown in Fig. 4, and is characterized by the following parameters: (1) the notch length at the specimen surface, a, ; (2) the specimen width, W, and (3) the thickness, B. A total of 60 specimens were tested with approximately 10 in each category. In both specimen configurations, the tests were carried out in an Instron load frame with a crosshead speed of 0.02 in. min-‘. Due to the low-carrying capacity of both the notched specimens, a full-scale load of 50 lbs was selected for the test. Fracture toughnesses for the straight- and chevron-notched specimens were determined according to Munz et al. (1980a, b).
0
-------A Fig. 4. Plan view of the general chevron-notched fracture toughness sample with nomenclature. 3. RESULTS AND DISCUSSION
The flexural test is commonly employed in strength characterization of ceramic matrix composites due to the problem of grippping and aligning the specimen during the tensile test. The complexity in performing a high-temperature test with tensile fixture and the cost of tensile specimens are major concerns in characterizing ceramic matrix composites, but the common practice of performing a bend test appears to give a higher strength than the data obtained from a standard tensile coupon for both polymer and ceramic matrix composites (Whitney et al., 1982; Lamicq et al., 1986). Flaws existing in a specimen are subjected to a uniform maximum stress field during tensile loading. However, during flexure, the stress field varies almost linearly from the neutral axis of the beam cross-section. This variation in the stress field reduces the possibility of occurrence of flaws within the maximum stress field, and hence gives a higher strength. Again in bend tests, three-point loading produces a higher strength as compared to four-point loading. This is because the smaller volume under maximum stress gives a higher local strength. This difference in strength suggests the importance of finding the ratios for characteristic flexural strength to tensile strength using Weibull statistical strength theory for brittle materials (Weibull, 1939). Since Weibull statistics contain a volume term, they can predict the effect of volume on the strength parameter. The fracture toughness Ki, , measured from the critical stress-intensity factor K,, , is applied to linear elastic fracture mechanics. Ceramic materials behave in a linear elastic manner at least at low temperature, which reasonably justifies the measurement of the fracture toughness through the K1, value. The four-point bend method with both straightand chevron-notched specimens appears to be a promising candidate for standard K,,
242
S. SALEIUN
Flexurul
et al.
strength
( MPa 1
Fig. 5. Weibull plot of flexural strength data.
measurement techniques. Despite such a presumption, the results show a discrepancy in the measured values due to the influence of the specimen dimensions, loading rates and loading fixtures, showing the need for a standard testing technique in order to obtain a true K,, value for ceramic materials. Data for the flexural strength of A&O,, SiC,/Al,O, and SiC,/A1203 were analyzed using the Weibull distribution curves shown in Fig. 5, to establish the probability of failure occurrence. It is evident from Fig. 5 that the flexural strength of single-phase monolithic A1203 has been increased almost one and half times due to the incorporation of Sic whiskers, whereas in the case of Sic platelets the strength of the composite has been decreased. Most of the flexural strength data in Fig. 5 show linearity and a well-distributed fit in the Weibull plot, except for few in the case of A120,. The bad fit represents the material behavior and the quality of the data. The data marked with an asterisk for A120, show a material defect due to a significantly lower flexural strength. The value of the shape factor parameter of monolithic A&O3 would be higher if these data were omitted from the plot. A significantly higher value for the shape factor parameter denotes less scatter in the data of the flexural strength of SiC,/A1203. The difficulty with the hot-pressing technique as regards the orientation of the platelets is easily recognized from the SEM micrographs shown in Figs 6-8. To obtain the optimum strength properties in platelet-reinforced composites, flatness and freedom from surface and edge cracks are essential, but a high degree of parallelism in platelet distribution is more important than the uniformity of the particles themselves. The lack of such a parallelism in the platelet distribution is revealed by the SEM micrographs. Also, whenever a multiphase material is processed at elevated temperatures, the differences in the thermal expansion coefficients and elastic modulii of the phases result in residual stresses upon cooling to room temperature. These residual thermomechanical stresses have always been of interest from both the strengthening and the weakening perspective. The geometry of the inclusion, expressed as the aspect ratio (length/diameter or Nd) directly affects the magnitude of this internal stresses (Zhuang and Bradt, 1989). This ratio is a highly significant factor which can either amplify or reduce the thermal and elastic effects on the internal stresses. When it increases, the stress inside the inclusion parallel to the 1 axis significantly increases, and the tangential stress parallel to the I axis at the interface of the inclusion decreases. As a result, an inclusion with a whisper shape will carry more unidirectional load than one with a spherical shape. The results of the present study show that composites reinforced with platelets have a very low flexural strength and the low I/d ratio of the platelets is a significant factor in causing higher residual micromechanical stresses. However, an experimental investigation of such
Effect of reinforcement geometry on composites
Fig. 6. Scanning electron micrograph of fractured surface of SiC,/Al,O, .
243
Fig. 7. Scanning electron micrograph of fractured surface of SiC,/Al,O, (straight-notched specimen).
Fig. 8. Scanning electron micrograph of fractured surface of SiC,/Al,O, (chevron-notched specimen).
Effect of reinforcement geometry on composites
245
residual stresses is necessary before drawing any conclusions. The porosity due to incomplete densification is also of critical concern in platelet composites. These defects may be due to improper platelet/powder blending or because of harvesting of Sic platelets prior to mixing with Al,O, . Figures 9 and 10 show the Weibull distribution of the fracture toughness of monolithic A&O,, SiC,/A120, and SiC,/Al,O, composites obtained using the four-point bend test for straight- and chevron-notched specimens, respectively. A comparison of the Weibull characteristic values with the average value for both the flexural strength and facture toughness is presented in Table 1. It is observed that the K,, of A&O3 is increased with the incorporation of Sic whiskers and has a maximum value of 5.09 MPa 6 for straight-notched specimens and 7.69 for chevron-notched specimens. This observed Table 1. Comparison
of Weibull
characteristic lives and average vdues of flexural strength and fracture toughness
Flexural strength 4-point bend (MM
Materials
AW, Sic, /Al,O, Sic, /A&O,
Weibull characteristic life 391.46 620.72 122.83
Fracture toughness 4-point bend’ (MPa Ml”)
Average
Weibull characteristic life
370.22 608.59 119.85
4.27 5.17 3.53
Fracture toughness 4-point bend’ (MPa M”‘)
Average
Weibull characteristic life
Average
4.23 5.09 3.21
5.97 7.86 4.83
5.63 7.69 4.65
‘Straight notch. r Chevron notch.
increase is due to changes in the microstructure. It is observed that the average values are lower than the Weibull characteristic values. It is also evident that the chevron-notched specimens show higher K,, values than those of the streight-notched specimens for both Sic, and Sic, composites. The addition of whiskers to monolithic A&O3 results in crack deflection, whisker pullout and possibly crack bridging, as evident from the SEM micrographs (Fig. 6) of the fractured surface of the SiC,/A1203 specimen. Whisker pullout is detected in both the fractured surfaces. Holes are observed where whiskers dislodged prior to fracture, and in some cases whiskers are observed to be protruding above the fractured surface.
K,,
(MPaM’?
Fig. 9. Weibull plot of fracture toughness data (straight-notched COE 2:4-B
specimen).
246
2
J IO
Fig. 10. Weibull plot of fracture toughness data (chevron-notched specimen).
The state of stress at the matrix/whisker interface plays an important role in the toughening mechanisms. Due to the thermal coefficient expansion mismatch of A1203 and Sic,, large compressive stresses are developed in the Sic whiskers and tensile stresses are developed in the Al,O, matrix. Since the whiskers are in compression and the matrix is in tension at the interface, the main propagating crack is attracted to the tensile region and is deflected along the interface. It is clear from Figs 9 and 10 that the fracture toughness values of the SiC,/A1z03 composites are significantly lower than that of monolithic A&O,. The reasons given for the discrepancy in flexural strengths are also applicable to the fracture toughness. The SEM micrographs of platelet composites (Figs 7 and 8) show less evidence of any commonly implied toughening mechanisms observed in whisker composites. Analytical investigations have shown that the degree of toughening depends on the shape and therefore on the surface area of the dispersoid (Shalek et al., 1986). Both analytical and experimental investigations of the failure mechanisms with different sizes of platelets are necessary for a complete understanding of the effects of such reinforcements. 4. CONCLUSIONS
From the results of the investigation (1)
(2) (3)
the following conclusions can be drawn.
The flexural strength and fracture toughness of monolithic A1z03 increase under the incorporation of Sic whiskers. In the case of platelets both the strength and toughness of the composite appear to decrease, in comparison to monolithic AlzOs . It appears that chevron-notched specimens provide higher K,, values that those obtained using straight-notched specimens, for both Sic, and Sic, composites. Scanning electron microscopy reveals that there is lack of parallelism in platelet orientation which necessitates the development of a technique to control the platelet orientation in processing.
Acknowledgements-The authors would like to acknowledge the support from the Office of Naval Research through contract number NO0 14-86-k-0765 and the United States Air Force, Office of Scientific Research through contract number F-49620-20-89-C-0016DEF for this research. REFERENCES Becher, P. F., Hsueh, P. and Tiegs, T. N. (1988). Toughening behavior in whisker-reinforced ceramic matrix composite. J. Am. Ceram. Sot. 71(12), 1056-1061. Bulijan, S. T., Baldani, J. G. and Huckabee, M. L. (1987). Si,N,/SiC composites. Am. Ceram. Sot. Bull. t%(2), 347-352.
Effect of reinforcement geometry on composites
241
Bulijan, S. T. and Sarin, V. K. (1987). Silicon nitride-based composites. Composites M(2). 99-106. Claussen, N. (1976). Fracture toughness of A&O, with an unstabilized ZrO, dispersed phase. J. Am. Cer. Sot. 59(1-2), 49-51. Homeny, J., Vaughn, W. L. and Ferber, M. K. (1987). Processing and mechanical properties of SiCWhisker-A&O,-matrix composites. Am. Ceram. Sot. Bull. 66(2), 333-338. Jenkins, M. G., Kobayashi, A. S., White, K. W. and Bradt, R. C. (1987). Crack initiation and arrest in a Sic whisker/Al,O, matrix composite. J. Am. Ceram. Sot. 70(6), 363-395. Lamicq, P. J., Bernhart, G. A., Dauchier, M. M. and Mace, J. G. (1986). Sic/Sic composite ceramics. Am. Ceramic Sot. Bull. 65(2), 336-338. Marshall, D. B. and Ritter, J. E. (1987). Reliability of advanced structural ceramics and ceramic matrix composites-A review. Ceram. Bull. 66(2), 309-317. Munz, D., Busbey, R. T. and Shannon, J. L. (1980a). Fracture toughness determination of alumina using fourpoint bend specimens with straight-through and chevron notches. J. Am. Ceram. Sot. 63(5-6). Munz, D. J., Shannon, J. L. and Busbey, R. T. (1980b). Fracture toughness calculation from maximum load in four-point bend tests of chevron notch specimens. Int. J. Fracture 16R, 137-141. Neegaard, L. J. and Homeny, J. (1989). Mechanical properties of beta-silicon nitride whisker/silicon nitride matrix composites. Technical Report UIUC-NCCMR-89-0008, Department of Material Science and Engineering, University of Illinois, Urbana-Champaign. Shalek, P. D., Petrovic, J. J., Hurley, G. F. and Gac, F. D. (1986). Hot-pressed Sic whisker/Si,N, matrix composites. Am. Ceram. Sot. Bull 65(2), 351-356. Sing, J. P., Smith, S. and Scattergood, R. 0. (1989). Observation on the toughening of Also,-SiC composites. Materials and Components Technology Division, Argonne National Laboratory, Argonne, IL. Tiegs, T. N. and Becher, P. F. (1985). Whisker-reinforced ceramic composites. Am. Ceram. Sot. Bull. 64(2), 298-304. Wei, G. C. and Becher, P. F. (1985). Development of Sic whisker reinforced ceramics. Am. Ceram. Sot. Bull. 65(2), 298-304. Weibull, W. (1939). A statistical theory of the strength of materials. Ing. Vetenskaps Akad., Hand/., No. 151, pp. 5-45. Whitney, J. M., Daniel, I. M. and Pipes, B. (1982). Experimental Mechanics of Fiber Reinforced Composite Materials. Society for Experimental Stress Analysis, Monograph No. 4. Zhuang, L. and Bradt, R. C. (1989). Micromechanical stresses in Sic-reinforced A&O, composites. J. Am. Ceram. Sot. 72(l), 70-77. Zweben, C., Smith, W. S. and Wardle, M. W. (1979). Test methods for fiber tensile strength, composite flexure modulus, and properties of fabric-reinforced laminates. In Composite Materials: Testing and Deisign (Fifth Conf.), ASTM STP 674, pp. 228-262. ASTM, Philadelphia, PA.