surements were expressed in the results. Wear scar was Starting ceramic powder compositions of samples(vol % examined by scanning electron microscope equipped with T20 an energy dispersive X-ray analyzer 995 79.5 AKP-30, Sumitomo Chemical, Osaka, Japan EP, Junsei Chemical, Tokyo, Japan The density of sample was higher than 99% of theoreti SCW#1, Tateho Chemical Industries, Hyogo, Japan. cal density. Fig. I shows friction coefficient of samples The friction coefficient of sample A against the silicon nitride ball is 0.7 which was lower than that of a sample and w20 were prepared according to the same procedure with 10 vol. SiC whiskers. Friction coefficients of sam- for sample A except for the fact that planetary ball ples with 20 vol. whiskers were lower than the corre- milling without whisker was stopped after 7.5 h and then sponding values of samples with 10 vol. whiskers. The esumed for 0.5 h after adding the whiskers. Hot pressing measured friction coefficient was sensitive to the contact was performed at 2123 K under 30 MPa for I h after that. surface roughness. Fig. 2 shows worn surface of sample A Diameter and length of the whiskers were 1-1. 4 and and W10 Worn area of samples were smooth and showed 10-20 um, respectively according to information supplied plastic deformation. This is thought to be due to the fact from the manufacturer. For samples T10 and T20, a modi- that the sample was at a very early stage of wear and the fied tape casting method was employed for aligning the mild wear regime prevailed. Sample A has a smoother whiskers. Tape casting slurry was prepared by using plane- wear track than the other, which was consistent with tary ball milling. Metyl-Isobutyl-Ketone, KD-1(ICI Chem- measured friction coefficients. Some grain boundaries of ical, Barcelona, Spain), polyvinyl-butyral (Aldrich Chemi- the samples showed up as a result of the wear test, which cal, New York, USA)and di-butyl phathalate(Aldrich contributed to increasing roughness of the contact surface Chemical) were used as solvent, dispersant, binder andand variations in the friction coefficient. Grain size of plasticizer, respectively. Care was exercised to minimize sample A was smaller than that of sample w10 due to damage of the whiskers during mixing, mixing without the lower hot pressing temperature. When the sample con- whiskers was carried out for 7.5 h and then for 0.5 h after tained Sic whiskers harder than the silicon nitride ball. the whisker addition. The slurry was poured into a reservoir of whiskers scratched the surface of the ball and ridges tape casting equipment that was modified for better align- formed starting from the whisker on the surface of the ment of the whiskers. Details of the modification was composite sample as shown in Fig. 2(b) Reported knoop described elsewhere [6]. Tape cast products were cut and hardness values of single crystal SiC for(0001) plane in laminated at 353 K under 50 MPa for 0.5 h. Binder [1010] and [1120] direction were 28.6 and 28.98 GPa, burn-out was performed at 823 K for 10 h in open air, and respectively [8]. The values for (1010) plane in [0001] and then the sample was hot pressed at 2123 K under 30 MPa [1120] direction(parallel to c-axis and normal to c-axis on Tribological behavior was studied by using the same ball-on-reciprocating flat geometry that was previously reported [7]. Silicon nitride ball (NDB100, Norton, North surface normal to hot pressing direction(conventional H P. boro, MA, USA)of 6.35 mm in diameter was used against normal to tape casting direction(tape casting flat alumina-SiC whisker composite sample surfaces which surface normal to lamination direction(tape casting were polished with I um diamond slurry. Normal load was 40N and average speed was 10 mm/s. Reciprocating stroke was 5.64 mm and duration was I h. Wear tests were performed at 403 K in order to eliminate moisture ad sorbed on the surface. In order to examine the effect of whisker orientation on wear rate. tests were carried out on two surfaces(tape surface and surface normal to tape casting direction) for samples W10 and W20, and in two o=9= directions on the tape surface for samples(parallel and normal to tape casting direction) T10 and Wear volume of the flat sample was obtained by wear track length times cross-sectional worn area that was measured from the center portion of the groove by a surface pro- Whisker content(vol % ometer after cleaning in a ultrasonic bath. Each wear test Fig. 1. Friction coefficients between the alumina-Sic whisker composite was repeated three times and average values of the mea- and the silicon nitride ballD.-S. Lim et al.rWear 225–229 1999 868–873 ( ) 869 Table 1 Starting ceramic powder compositions of samples vol.% Ž . Sample A W10 W20 T10 T20 a Al O 99.5 89.5 79.5 89.5 79.5 2 3 MgO 0.5 0.5 0.5 0.5 0.5 b c SiC whisker – 10 20 10 20 a AKP-30, Sumitomo Chemical, Osaka, Japan. b EP, Junsei Chemical, Tokyo, Japan. c SCWa1, Tateho Chemical Industries, Hyogo, Japan. and W20 were prepared according to the same procedure as for sample A except for the fact that planetary ball milling without whisker was stopped after 7.5 h and then resumed for 0.5 h after adding the whiskers. Hot pressing was performed at 2123 K under 30 MPa for 1 h after that. Diameter and length of the whiskers were 1–1.4 and 10–20 mm, respectively according to information supplied from the manufacturer. For samples T10 and T20, a modi￾fied tape casting method was employed for aligning the whiskers. Tape casting slurry was prepared by using plane￾tary ball milling. Metyl-Isobutyl-Ketone, KD-1 ICI Chem- Ž ical, Barcelona, Spain , polyvinyl-butyral Aldrich Chemi- . Ž cal, New York, USA and di-butyl phathalate Aldrich . Ž Chemical were used as solvent, dispersant, binder and . plasticizer, respectively. Care was exercised to minimize damage of the whiskers during mixing; mixing without the whiskers was carried out for 7.5 h and then for 0.5 h after whisker addition. The slurry was poured into a reservoir of tape casting equipment that was modified for better align￾ment of the whiskers. Details of the modification was described elsewhere 6 . Tape cast products were cut and w x laminated at 353 K under 50 MPa for 0.5 h. Binder burn-out was performed at 823 K for 10 h in open air, and then the sample was hot pressed at 2123 K under 30 MPa for 1 h. Tribological behavior was studied by using the same ball-on-reciprocating flat geometry that was previously reported 7 . Silicon nitride ball NDB100, Norton, North- w x Ž boro, MA, USA of 6.35 mm in diameter was used against . flat alumina–SiC whisker composite sample surfaces which were polished with 1 mm diamond slurry. Normal load was 40 N and average speed was 10 mmrs. Reciprocating stroke was 5.64 mm and duration was 1 h. Wear tests were performed at 403 K in order to eliminate moisture ad￾sorbed on the surface. In order to examine the effect of whisker orientation on wear rate, tests were carried out on two surfaces tape surface and surface normal to tape Ž casting direction for samples W10 and W20, and in two . directions on the tape surface for samples parallel and Ž normal to tape casting direction T10 and T20. Wear . volume of the flat sample was obtained by wear track length times cross-sectional worn area that was measured from the center portion of the groove by a surface pro￾filometer after cleaning in a ultrasonic bath. Each wear test was repeated three times and average values of the mea￾surements were expressed in the results. Wear scar was examined by scanning electron microscope equipped with an energy dispersive X-ray analyzer. 3. Results and discussions The density of sample was higher than 99% of theoreti￾cal density. Fig. 1 shows friction coefficient of samples. The friction coefficient of sample A against the silicon nitride ball is 0.7 which was lower than that of a sample with 10 vol.% SiC whiskers. Friction coefficients of sam￾ples with 20 vol.% whiskers were lower than the corre￾sponding values of samples with 10 vol.% whiskers. The measured friction coefficient was sensitive to the contact surface roughness. Fig. 2 shows worn surface of sample A and W10. Worn area of samples were smooth and showed plastic deformation. This is thought to be due to the fact that the sample was at a very early stage of wear and the mild wear regime prevailed. Sample A has a smoother wear track than the other, which was consistent with measured friction coefficients. Some grain boundaries of the samples showed up as a result of the wear test, which contributed to increasing roughness of the contact surface and variations in the friction coefficient. Grain size of sample A was smaller than that of sample W10 due to lower hot pressing temperature. When the sample con￾tained SiC whiskers harder than the silicon nitride ball, the whiskers scratched the surface of the ball and ridges formed starting from the whisker on the surface of the composite sample as shown in Fig. 2 b . Reported knoop Ž . hardness values of single crystal SiC for 0001 plane in Ž . wx wx 1010 and 1120 direction were 28.6 and 28.98 GPa, respectively 8 . The values for 1010 plane in 0001 and w x Ž . w x w x 1120 direction parallel to Ž c-axis and normal to c-axis on Fig. 1. Friction coefficients between the alumina–SiC whisker composite and the silicon nitride ball