No abstract

A bending technique for the strength measurement of glass fibers is described and an analysis is presented which determines the effective tested length as a function of the statistical parameters which describe the fracture properties of the fiber. The analysis is used to compare strength data determined in both tension and bending for various representative fibers. It is found that the tensile strength cannot be predicted from the strength in bending and vice versa because the tested lengths differ by at least 3 orders of magnitude. Thus, while bending does not replace tension as a measurement technique, it does provide additional valuable information about the flaw size distribution.

No abstract

No abstract