Search

Slow crack growth parameters for SiC-based filaments. τ is the lifetime, σ app the stress, n f the stress exponent and Ea τ the apparent activation energy describing the Arrhenius evolution of the environment constant A with temperature. [Display omitted] • Modified time to failure prediction expression applied to various SiC-based fibers. • Stress exponent subjected to a scale effect, induced by fiber misalignment. • Estimates for SCG parameters of SiC-based filaments. • Stress-probability time diagrams construction from simulation output. • Correlation between stress exponent and chemical composition (O and H content). The delayed failure of SiC fibrous reinforcement has continuously been investigated to warrant the long term performances of Ceramic Matrix Composite (CMC). Chiefly assessed on multifilament tow samples to alleviate some handling difficulties, subcritical crack growth (SCG) parameters are however ruled by structural artifacts which hinder the identification of intrinsic filament behavior. In this paper, we propose to estimate the true filament parameters for 5 fiber types from bundle behavior using a recently communicated Monte Carlo algorithm integrating flaw and stress distributions through a deterministic fracture mechanics law under Paris' formulation. So computed tow lifetime are broadly dispersed, encompassing raw data, and show a structure-dependent scale effect, revealed by n filament >n tow where n is the stress exponent. The relationship between SCG coefficient and chemical composition of the substrate is discussed and highlights the major effect of doping elements (Ti or Zr), oxygen or hydrogen content. [ABSTRACT FROM AUTHOR]

Abstract: Silicon nitride is used for demanding tasks due to its high stiffness, strength and, especially, its high fracture toughness. Examples include cutting tools, forming rolls and ball bearings. The microstructure is characterized by elongated β-Si3N4 grains of different size and shape, which lead to the increased fracture toughness. Consequently, the paper will present an algorithm for the generation of three-dimensional and periodic silicon nitride-like microstructures, which will be used for micromechanical finite element simulations. The structure generation algorithm enhances the sequential adsorption technique with growth of particles and steric hindrance, which are motivated by experimental results. Results of the structure generator, such as the pseudo-time evolution and its statistical geometric distributions are presented and compared to literature data. With the finite element simulations, using a periodic unit cell, a validation of the model with literature values for Young''s modulus and Poisson''s ratio was possible. [Copyright &y& Elsevier]