Statistics and Microphysics of the Fracture of Glass

TL;DR

This paper explores the fracture behavior of glass, applying statistical models to understand failure sites and predict strength variability, with implications for assessing the intrinsic strength of silica fibers.

Contribution

It introduces the use of classical weakest link models to analyze glass fracture, proposing methods to determine failure site distribution from skewness and dispersion data.

Findings

Calculated skewnesses for failure strength distributions

Potential to identify number of failure sites from statistical measurements

Framework for comparing future experimental data

Abstract

The tensile strength of fused silica fibers is believed to approach its intrinsic value at low temperature, and modern experiments indicate very small, perhaps unmeasured, intrinsic dispersion in this strength. I consider the application of classical ``weakest link'' models to this problem in an attempt to determine the number and therefore the nature of the failure sites. If the skewness as well as the dispersion (Weibull modulus) of failure strengths are measured it may be possible to determine both the number of sites and the distribution of their strengths. Extant data are not sufficient, but I present calculated skewnesses for comparison with future data.