SiC whisker composites fabricated by two techniques

MATERIAlS

SCIEHE& E_IMEERIIiG ELSEVIER

Materials Science and Engineering A209 (1996) II J 115

A

High-temperature deformation of Zr0 2 - A1 2 0 3 /SiC whisker composites fabricated by two techniques J.M. Calderon Moreno'\ A.R. DeArellano-Lopez't, A. Dominguez-Rodrigueza , J.L. Routbort h "Department Materia Condensada & Institl/to de Cienciils de Milteriilles de Serif/a, Unirenidad de Serilla-CSIC, SetJ//a, Spain bEnaKl' Techn%K!' f)it'ision, ArKonne !Viltiona/ Lahori/tor!', ArKonne, IL 60439, USA

Abstract Zr0 2 -AI 20,/Sie whisker-reinforced composites, with whisker volume fractions of 0'1"-28'1,,, fabricated both by powder and precursor processing, have been deformed at temperatures of 1300--1500 °e under constant compression rates of 1.7 x 10 'to 6.8 x 10 - 5 S - I. Above 1400 0c, a stress at which the work-hardening rate became zero could be measured and correlated with whisker content. On the contrary, at 1300 °e all samples broke within the elastic regime. At 1350 °e, increased whisker content appeared to inhibit fracture, so plastic behavior was obtained for samples containing 28% SiC. In the range of temperatures and compression rates, noted above. stress exponents were determined and tentatively correlated with microstructural features and their evolution during plastic deformation. Ke)'ll'Ords: Powder processing; Precursor processing; Plastic deformation

1. Introduction Fabrication and mechanical properties of the system AI 2 0 3 -Zr0 2 • known as ZTA. have been of interest in the last few years [1,2]. ZTA shows enhanced fracture toughness (K ,c = 10 MPa m 1/2 ) at room temperature [3] and is a promising material for high-temperature applications [4]. although its strength is reduced because of microcracking, one of the micromechanical toughening mechanisms [5]. Most results on ZTA room-temperature mechanical properties show a strong dependence on the volume fraction of each component, the ratio of tetragonal to monoclinic Zr0 2 • and ZrO:, grain size [2]. In parallel. the addition of ceramic whiskers as reinforcements for a variety of ceramic matrixes has shown interesting mechanical properties at both room and elevated temperatures [6]. Some of the now classical systems, such as AI 20 3 -SiC w , have been widely studied [7,8]. The enhanced room-temperature fracture properties (both toughness and strength) of whisker-reinforced composites were explained by the combined action of three toughening mechanisms, debonding, bridging, and pull-out [9], which resulted in a maximum K /c of 8 0921-5093/96/$15.00

I!,,)

1996

SSDI 0921-5093(95)10144-6

Elsevier Science SA All rights reserved

MPa m I!'. The recent literature has improved our understanding of the role of the whiskers in the reporting of reduced creep rates measured in samples containing these stiff filaments [10-- 12]. Partial inhibition of grainboundary sliding (GBS), the main creep mechanism in fined-grained ceramics [13], is considered responsible for the enhanced creep resistance. For high stresses and compression rates, damage is produced in the vicinity of whiskers because they act as stress concentration sites, and diffusion cannot accommodate strain. The whiskers form a network that prevents brittle fracture of the samples. The possibility of improving fracture toughness and fracture strength made the study of ZTA-SiC w composites [14] particularly interesting. Initial results indicated that a fracture toughness of up to 11 MPa m 12 could be achieved while retaining significant strength [3]. Examination of the high-temperature behavior of this system naturally follows as a research effort. The aims of this study are to establish the temperature ranges of creep and provide a preliminary view of the microstructural role of Zr0 2 and SiC whiskers in creep mechanisms.

112

J. M. Calder
2. Materials Two kinds of ZTA-based SiC whisker composites, both containing 5.5 vol.'!,() Zr0 2 , have been studied. Additionally, they contained 0, 10, and 28 vol.% SiC whiskers. In one case, the samples were processed by wet mixing raw powders (pd-samples). The AI 2 0 3 was in the form of ct-A1 2 0 3 and the Zr0 2 powder was in the form of monoclinic high-purity m-Zr0 2 with a typical grain size of 0.5 ,um. In a second case, processing was carried out by a sol-gel technique with an Al 1 0 3 solution and a Zr0 2 precursor (pc-samples). The slurry was calcined, screened and dried. This latter procedure should result in better mixing of the components. In both cases, SiC whiskers were added in the correct proportion. Densification was achieved by uniaxial hotpressing of the green bodies at 1600 dc. The name of each specimen includes a number that specifies the volume fraction of SiC. Full density was obtained for all composites except pc-ZTA28 (93%). Microstructure was studied by SEM and TEM, with computer-aided image analysis. Polished surfaces of as-received and deformed specimens, thermally etched in air at 1450 DC, revealed grain morphologies of the type shown in Fig. I. TEM samples were prepared by standard techniques (dimpling, ion-beam milling, etc.), and observations included grains, whiskers, disloca-

Table I Grain Size (fim) of ZTA- SiC w Composites

o SiC

w

voL';;', Powder proc. (pd-) Precursor proc. (pc-)

4.7

2.6

10 SiC w voL% 1.9 1.2

28 SiC w voL% 1.1 1.3

tions, pores, cracks, and glassy phases. Some features are shown in Fig. 2. More details on microstructural features of ZTA ceramics are provided in Ref. [15]. Grain sizes of the as-fabricated samples are included in Table I.

3. Mechanical tests All materials were deformed in an Instron 1185 machine, in air, at 1300-1500 °C under uniaxial compressive constant crosshead speeds; initial strain rates were 1. 7 x 10 5 to 6.8 x 10 - 5 S - 1. Specimen sizes were approximately 3 x 3 x 6 mm. To establish the sensitivity of the strain rate to stress, changes in compression rates were made at constant temperature, and stress was measured when the workhardening rate reached zero. Note that a zero workhardening stress does not mean a deformation steady state. The latter is appropriate only when microstructure is unchanged. The stress exponents (n) were calculated from the creep equation t = A (1'/1 where t is the strain rate, (1' is the stress, and A is a parameter that depends principally on temperature and grain size [16].

4. Results and discussion

Fig. 1. SEM image of etched pd-ZTAO, showing grain morphology.

Typical stress vs. strain plots are shown in Fig. 3 (a) and (b). A summary of results in terms of maximum stresses before fracture and zero work-hardening stresses follows. Maximum stresses between 1300 and 1500 DC, at 1. 7 x 10- 5 S - 1 compressive strain rate, for all specimens, are detailed in Table 2, which also indicates whether fracture (frac.), fracture after some softening (mix.) or no fracture with measurable flow stress (flow) was observed. At the lowest temperature, all specimens fractured in the elastic regime. At 1350 DC, samples containing < 100/.) whiskers fractured after exhibiting some plastic flow regime, while samples with 28'% whiskers behaved plastically. For the pd-processed materials, the addition of whiskers did not significantly change the maximum stress at a given temperature. However, at 1300 and 1350 DC, there was increase of > 100 MPa from pc-ZTAO to pc-ZTA IO. As stated before, pc-ZTA28 materials

J.M. Calder/Jn Moreno

el

al. Materials Science and Engineerinf!, A209 (1996) 111-115

have high porosity (7%), which accounts for their poorer mechanical response. However, this result must be combined with the observation of a larger reduction of grain size in pd-composites than in pc-composites owing to the addition of whiskers. Previous results showed that the composites fabricated from powders have larger AI 20 3 grain sizes and larger Zr0 2 particle sizes than composites fabricated from precursors [IS]. Several authors have demonstrated that larger Zr0 2 particles result in reduced strength due to microcracking [2], consistent with our observations.

F: 573 MPa

CSR = 1,7 x10· pc-ZTA10

5

Table 2 Maximum stresses at 1.7 x 10- 5

1300

1350

S·1

1400

500

-A--

• •

~MPa /{!,

400

Ii

a.

~

~ .:>~

o

w

g:

s:Y

en

Y

13500 C 14'iOoC

--,6.

/y/

~300

13000 C

14000 C

1450

15000 C

Brittle fracture below 1400 0 C

200

1500

143 MPa

--i.-._....-.--.--.-i.-...-.--.--A--. 100

65 MPa

•••••••• 3

21

6

(a)

STRAIN{'!o)

pd-ZTA28 CSR=1.7 x10- 6 S-1

---A-

300

.-•

Ii 0..

~

•

F 205 MPa

~200

g:w en

S-l

113

compressive strain rate

Processing

SiC vol.'Yr,

Stress (MPa)

pdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpcpdpc-

0 0 10 10 28 28 0 0 \0 10 28 28 0 0 10 10 28 28 0 0 10 10 28 28 10 10 28 28

385 434 404 573 394 290 231 296 230 400 205 163 95 157 86 143 84 75 51 54 50 65 48 32 36 40 32 18

frac. frac. frac. frac. frac. frac. frac. frac. mix. mix. flow flow flow flow flow flow flow flow flow flow flow flow flow flow flow flow flow flow

The observation that addition of 28 vol.% SiC suppresses brittle fracture at 1350 °C in the pc-material is an indication that whisker reinforcement inhibits microcracking. The inhibition of brittle fracture is particularly effective with high whisker loadings. Some authors claim that microstructures containing whiskers are more damage-tolerant [11,17].

13000 C 1,350o C

pd-ZTA10 CSR= 1.7(1)-3.4(2)-6.8(3) x10

14000 C

120

---A-.

1400 0 C

15000 C

-..

1460 0 C

100

•

Ii

~

100

a:: ..... en

84MP

·1

/Ii.

0

84 MPa n= 2 (1-2)

1600 C

f

60

2~.2) t

'.. . . . . . . . . . . .

r.

........ t+ 89 MPa

.......... , •••••••••••

52 MPa n=

11 ~ MPa

~

69 MPa n= 2.7 (2-3)

80

40 1

48 MPa

S

~~~
14500 C

en en w

-5

..

47MPa

37 MPa n= 3 (1-2)

20

o (b)

3

6

9

12

15

18

STRAIN ('!o)

0 .......L~---l_L-..L-....L~---l_L-..L-....L--I...---I_.L......L--' 16 12 8 o 4 STRAIN('!o)

Fig. 3. Stress vs. strain plot of mechanical tests for (a) pc-ZTAIO and (b) pd-ZTA28 showing maximum stresses before fracture and zero work-hardening stresses.

Fig. 4. Stress vs. strain plot of mechanical tests for stress exponent determination in pd-ZTAIO.

114

J.M. Calderon Moreno et al. / Materials Science and EnKineering .1209 (/996) / / / - / /5

Table 3 Stress exponents at or above 1400 °C Temp. (OC)

Process

1400

pdpcpdpcpdpcpdpcpdpcpdpc-

1450

1500

SiC vol. '/,:,

n

10

2

10

1.7

28 28 10 10 28 28 10 10 28 28

2.5 2.1 2.6 2.2 4.5 2.9 3 2.5 3.5 3

Above 1400 °C, all composItions deformed plastically, and flow stresses (with extended zero work-hardening regimes) that developed even up to 20% strain were determined for all samples. These stresses are also summarized in Table 2. Because of its poorer density, the pc-ZTA28 composite exhibited the lowest flow stresses at all temperatures. For a fixed fabrication process, whisker loading did not significantly affect flow stresses. However, one must consider that grain size, d, is normally reduced because of the addition of whiskers [10,12]. Models of creep mechanisms in which the transport of matter [16] predicts that the strain-rate dependence is proportional to l/dP , with p = 2 to 3, or that the stress is proportional to d P at a constant i:. Taking the values in Table I, one should expect that ZTA 10 samples would be 5 to 10 times softer than their matrices, and that pd-ZTA28 would be IS to 60 times softer than its matrix. Values in Table 2 show that the composite materials are stronger than expected. Finally, all other compositions based on the precursor-fabricated matrix showed higher flow stresses than the powder-fabricated samples, despite their smaller grain sizes. Some of the differences between the pc- and the pd-fabricated materials may be the result of intergranular glass phases. Analytical TEM is in progress. Such results indicate that precursor fabrication is a promising technique, and that some balance between whisker addition and grain-size reduction is needed to obtain the optimal mechanical properties. An example of experiments to calculate stress exponents is given in Fig. 4. The values of n are listed in Table 3 and are typically close to or higher than 2 (3 and 4 in some cases) and increase with increasing temperature. Results for creep mechanisms of ZTAO can be found in Ref. [15]. Several creep mechanisms might be responsible for high stress exponents, but most are related to dislocation activity [16]. Microstructural analysis (Fig. 5) revealed no detectable dislocation activity and that

damage is present as described in previous publications on ZTA and AI 20 3-SiC w [10, IS]. Damage is produced, but the whiskers prevent fracture from the damage. Zero work-hardening states of plastic deformation are then established and flow stresses are measured, but increases in compression rates produce smaller changes in flow stress as the result of damage production, thereby providing a possible explanation for the higher stress exponents.

5. Concluding remarks Literature on composites shows that microstructure is the key for obtaining good mechanical properties [18]. Because toughening mechanisms are multiple and synergistic, this is very important in a three-component system. In particular, there are many critical factors for each of the phases: AI 2 0 3 grain size, Zr0 2 particle size and distribution, whisker distribution and nature of whisker/matrix interfaces, etc. This study has identified the plasticity temperature range in ZTA-based composites in compression at strain rates of ~ 10 5 S - '. It has also shown that precursor-processed ZTA composites have higher flow stresses than those of powder-processed composites. A natural following step is to complete the characterization of the SiC composites, extending the range of experiments to lower stress and slower compression rates, in order to identify the role of diffusioncontrolled creep mechanisms.

Acknowledgements The authors thank Lori Leasky of MER Corporation, Phoenix, AZ, for supplying the samples. Work in Spain was supported by CICYT Project MAT91-0978 of the Spanish Ministerio de Educaci6n y Ciencia. JLR was supported by the U.S. Department of Energy, BES, Division of Materials Science, under Contract W-31109-Eng-38.

Fig. 5. TEM image of pd-ZTA28 after deformation at 1400 °C, showing severe damage (crack) indicated by arrows.

J.M. Ca/da"n Alor('/fo cl a/.

,\4'l/cria!.1 SciCllcc and f,'nginceflng A:!()<) (/996) 11/

References [I] N. Claussen, Microstructural design of ZrO,-toughened ceramics, in N. Claussen, M. Ruhle and A.H. Heuer (eds.). Scicllcc alld Techn%gy 0(2r0 2 11. Adranccs in Ceralllics. Vol. 12. Amcrican Ceramic Society. Columbus. Oil. 1984, p..'25. [2] DJ. Green, R.HJ. Hannick and M.V. Swain. FrallSfori//{/lion Toughening o( Ceramics. CRC Press. Boca Raton. IL. 1989. pp. 157 192. [,] N. Claussen. 1. Steeb and R.F. Pabst. Elkcts of induced microcracking on fracture toughness of ceramics. Alii. CcrWII. Soc. Bu//.. 56 (1977) 559. [4] R.C. Garvie, Microstructure and performancc of an aluminaZr0 2 tool-kit, J. Alata. Sci. LCII .• 3 (1984) "5. [5] S. Hori, M. Yoshimura and S. Somiya. Strength-toughness relations in sintered and isostatically hot-prcsscd ZrO, toughened AI 20" J. .1111. Cawn. Soc .. 69 (,) (1986) 169. [6] .I.R. Porter. Disperson processing of creep-resistant whisker-reinforced ceramic-matrix composites. M({/cr . .';ci. Ellg.. .1107 (1989) 127 Ll2. [7] P.F. Becher. G.e. Wei. Toughcning behavior in SiC '-whisker-reinforced alumina. J. .111I. Ceralll. Soc .. 67 (12) (19R4) C267 C269. [8] A.R. DeArellano-Lopez. F.L. Cumbrera. A. Dominguez-Rodriguez. K.C. Goretta and .I.L. Routbort. Compressive crcep of SiC-whisker-reinforced AI 20" J. Alii. Ccrall/. S·oc. 73 (5) (1990) 1297 ,00. [9] A.G. Evans. Perspectives on development of high-toughncss ceramics. J. .111I. Ceralll. Soc.. 73 (2) (1990) 187 206. [10] A.R. DeArellano-Lopez. A. Domingucz-Rodriguez. K.C.

[II]

[12]

[LI]

114]

[15]

[16] [17]

[18]

115

115

Goretla and 1. L. Routbort. Plastic deformation mechanisms in SiC-whisker-reinforced alumina composite. J. .111I. Ccram. Soc .. 76 (6) (199\) 425 4,2. P F Becher and T. N. Tiegs. Toughening behavior involving multiple mechansims: whisker reinforcement and Zro, toughening. J. .1111. Ccralll. Soc .. 7() (9) (1987) 657. H.-T. LIIl. P F. Becher. High-temperature creep deformation of alumina-SiC-whisker composites. J. Am. Ccralll. Soc.. 74 (8) ( 1991) 1886 189,. A.II. Heucr. NJ. Tieghe and R.M. Cannon, Plastic deformation of line-grained alumina (AU),): basal slip and unaccommodated grain boundary sliding. J. .111I. Cerall1. Soc .. 63 (I 2) (1980) 5.' 58. A.R. DcArellano-Lllpez. A. Dominguez-Rodriguez. K.C. Goretta and 1. L. Routbort. Plastic deformation of alumina reinforced with SiC-whiskers. Proc. Plastic Deforll1({/ion o( Ceralllics COli/.. SnOirhird. CT. Augusl 1994. in R.C. Bradt. Ch. A. Brookes and .I.L. Routbart (cds.). P/aslic Deformation 0/ Cawl1in. Plenum. New York. 1995. pp. 5"-542. 1. M Calderon-Moreno. A.R. DeArellano-L6pez. A. Dominguez-Rodriguez and 1.L. Routbort. Microstructure and creep properties of aiuminaZr0 2 ceramics. J. Eur. Ceramic Soc., 15 (1995) 98.' 988. W.R. Cannon and T.G. L.angdon. Review: Creep of ceramics, Part l: Mcchanical propertics. J. Maler. Sci.. 18 (198.') I 50. W. Gu . .I.R. Porter and T.G. Langdon. Evidence of anelastic recovery in silicon carbidc-whisker-reinforced alumina. J. Am. Ccram. Soc. 77 (6) (1984) 1679-1681. A.Cj. Evans. Structural reliability: a processing-depcndent phcnomcnon. J. Am. Ccram. Soc .. 650) (1982) 127.