Simulating the variability and scale effect for slow crack growth in Hi-Nicalon SiC-based tows: A parametric study

Abstract

The delayed failure time of SiC-based multifilament tows under static fatigue condition is as broadly scattered as for individual filaments, despite it is commonly used as strong scatter reduction. Moreover, the stress exponent (n) is hierarchy-dependent as revealed by a Monte-Carlo algorithm: decreasing from filament (micro) scale to tow (meso) scale. This is demonstrated to originate from the mismatch between the stress applied to the critical filament (affecting the growth kinetics), variable because of fiber misalignment in the tow, and its tow-averaged value used for endurance diagram construction. In the context of this algorithm, it is shown that n would evolve with fiber parallelism, tow stress range or the critical filament rank whereas filament strength distribution plays a secondary role. The tow structure shall therefore be considered as a major parameter for composite part design aiming long service life.