Page 125 - Fiber Fracture

110 A. Sayir and S.C. Farmer

pattern at a remote sensor plane. The diverse features of speckle patterns could then
be replayed in real time or correlated. The correlation was unconstrained as long as
the contrast was high. The level of contrast for the speckle pattern was in turn directly
related, among other factors, to the surface morphology of the material. A large percent
of the argon laser beam is internally reflected down the length of the sapphire, YAG
and Y2O3 fibers, thereby causing a weak signal return to the camera. The speckle
pattern therefore had a relatively low signal-to-noise ratio which, along with multiple
returns from the back surface of the fiber, contributed to the scatter in the data. The
weak contrast (dim speckles) increased the difficulty of tracking the speckle shift. With
each new reference acquired there was a small element of random error introduced
into the cumulative speckle shift record (Sayir et al., 1994). This random error further
increased scatter in the strain data. In order to reduce the uncertainty and increase
the correlation peak in the presence of an erratic pattern, a reference from a new
frame was obtained whenever the cumulative speckle displacement within the current
reference exceeded the maximum allowable speckle displacement per reference frame.
This effort, combined with careful alignment of the fiber, resulted in reduction of the
scatter to an acceptable level. Scanning electron microscopy (JOEL 840, UTW Kevex
EDS) and X-ray diffractometry (Phillips 3720XRD) were employed to characterize for
microstructures and phase identification.

RESULTS AND DISCUSSION

Fracture Characterktks of Single-Crystal Y203, Y3.415012 and A1203

The two single-crystal compositions which span the A1203-Y3A15012 region of
the AI2O3-Y203 phase diagram were selected for study. To advance our overall
understanding of the fracture characteristics of single-crystal fibers in the A1203-Y203
system, single-crystal (1 1 I) Y2O3 which exhibits different fracture characteristics was
also examined. Single crystals of A1203 and Y3Al5Olz were grown in (0001) and (1 1 I)
directions, respectively. The growth directions for these single crystals eliminate the
active slip systems (Snow and Heuer, 1973; Corman, 1991). The creep resistance of
single-crystal (OOO1) A1203 and (1 11) Y3A15012 are more than adequate for composite
applications up to 1500°C (Snow and Heuer, 1973). If single crystals of A1203 and
Y3Al5OI2 are loaded in the growth direction, the creep resistance of the material is
sufficiently large that there is no measurable contribution of creep the fracture process.
Efforts focused on the measurement of surface strain along one principal fiber axis,
determination of Young’s modulus, and analysis of the fracture characteristics of these
single crystals.
The micro-strain measurements of small-diameter fibers were performed using
Yamaguchi’s speckle-shift technique (Sayir et al., 1994). The average Young modulus
of (0001) A1203 and (111) Y3A15012 were 453 GPa (f36 GPa) and 290 GPa (f24
GPa), respectively. This average value for (Oool) A1203 (c-axis sapphire) is in good
agreement with reported data in the literature of 450 GPa (Wachtman, 1960). Yet, the
large scatter (8% standard deviation) in Young’s modulus of sapphire was the largest