Ingots having a cross-sectional area of 1.27 cm × 10.8 cm were directionally solidified for testing off-axis tensile properties. To avoid macroscopic banding due to inductive melting and thermal convection, growth rates of 2.2 cm/min were required. The resulting solidification structures were cellular.
The macroscopic tensile behavior of the eutectic alloy was similar to that observed for mechanically formed fiberreinforced composite materials; in particular, three distinct modes of fracture occurred. In Region I, where the angle(ϕ) between the growth direction and the tensile axis was small, the tensile strength of the composite was inversely proportional to cos2 ϕ. In Region II (20° ≤ ϕ ≤ 55°) and Region III (60° ≤ ϕ ≤ 90°), the tensile strength of the composite was inversely proportional to sin ϕ, cosϕ, and sin 2ϕ, respectively.
For all regions, failure was initiated by the fracture of large Al3Ni intercellular blades. In Region I, fracture was by a combination of tensile overload on a plane perpendicular to the tensile axis or shear on a plane inclined to the tensile axis at approximately 45°. In Region II, failure was by shear in a plane containing the growth direction and in a direction parallel to the fibers. For Region III, failure was by plane strain constraint with shear perpendicular to the longitudinal axis of the fibers.
The fracture of large intercellular Al3Ni blades can be explained by the variation of S13 about the tensile axis, and the corresponding intercellular tensile strains which result. In Region II, metallographic examination of tensile specimens shows a Poisson's ratio effect to be contributing to the strength of the composite; occasionally shear between adjacent cells is observed. By using the yield strengths of specimens having fibers oriented parallel to and perpendicular to the tensile axis, an approximate yield surface can be constructed. This yield surface accurately predicts the transition in failure mode from Region II to Region III.