Optimizing Ceramic Matrix Composite Interlaminar Fracture Toughness (Mode I) Wedge Test

Abstract

ASTM test standards for CMC’s Crack Growth Resistance (CGR) may exhibit a zig-zag (wavy) crack path pattern, and fiber bridging. The experimental parameters that may contribute to the difficulty can be summarized as: specimen width and thickness, interface coating thickness, mixed mode failure evolution, and interlaminar defects. Modes I crack growth resistances, GI were analytically determined at ambient temperature using wedge test, a modified double cantilever beam (DCB). Several Finite Element (FE) based Multi-scale modeling potential techniques were investigated: a) Multi-scale progressive failure analysis (MS-PFA); b) Virtual Crack Closure Technique (VCCT). Advantages and disadvantages of each were identified. The final modeling algorithm recommended was an integrated damage and fracture evolution methodology using combined MS-PFA and VCCT. The material tested in this study was a slurry-cast melt-infiltrated SiC/SiC composite with Tyranno ZMI fibers (Ube Industries, Kyoto, Japan) and a BN interphase. The fiber architecture consisted of eight plies of balanced 2-D woven five-harness satin. The total fiber volume fraction was about 30% with half of the fibers in the 0° direction and half in the 90° direction. All specimens had a nominal thickness of 4 mm. An alumina wedge with 18° head angle (2α) was used. In this method, a splitting force is created by inserting a vertically-moving wedge in a notch causing the arms to separate and forcing an interlaminar crack at the sharpest end of the notch The MS-PFA numerical model predicted the damage and fracture evolution and utilized the GENOA UMAT (User Material Subroutine) for Damage and FEM (Finite Element Model) stress intensity and LEFM (Linear elastic Fracture Model), Cohesive Model for Fracture. The analysis results (Fracture energy vs. crack length, Fracture energy vs. load, Fracture energy vs. crack opening displacement) matched the Mode I coupon tests and revealed the following key findings. Mode I-Wedge specimen exhibits: 1) failure mode is due to interlaminar tension (ILT) only in the interface section and a zig-zag pattern observed; 2) VCCT crack growth resistance is well matched to the test data; and 3) failure mode is a mixed mode behavior of Interlaminar tension (ILT) to interlaminar shear (ILS). The final Wedge test specimen configuration optimization includes the sensitivity of design parameters to CGR: a) wedge contact coefficient of friction; b) lever arms thickness, and c) inclined head angle, distance between the initial crack and wedge tip.