The Principles of Strength and Fatigue in Optical Fibers

Tensile strength is defined as the applied stress (tensile load per unit cross-sectional area) recorded at the instant of rupture for a test specimen. Historically, attempts to catalogue and report the intrinsic strength of glass were frustrated by considerable dispersion in the acquired measurement data. In addition to large variability, typical strength values were found to be one, two or even three orders of magnitude below that which would be expected theoretically. Unlike many intrinsic material constants for glass such as density, specific heat, or index of refraction, strength is statistical in nature. Both sample type and test condition have a bearing on the measured strength of glass. Specimen size, surface finish, and annealing contribute to the outcome of a measurement as well as the distinction between static versus dynamic, short-term versus long-term, or tensile versus flexural stress conditions. Strength of glass optical fibers is usually curtailed by the existence of particulates trapped in the glass or flaws (surface cracks) introduced during the manufacturing process and handling subsequent to manufacture. Such defects, which are submicroscopic in dimension, act as stress concentrators in the glass when fibers are placed under loads and therefore define weak points. The weakest site in a continuous length of fiber then determines its limiting strength. The strength of glass fiber may be influenced over time when held under stress in atmospheric air. Water vapor in the air migrates into surface flaws where a chemical reaction can occur. This reaction of water with the silicate network leads to fatigue of load-bearing elements in the glass by altering the shape of the crack tip. Total mechanical failure of the glass occurs when stress at the tip of a crack reaches the critical fracture stress. Practical applications of fiber optics involve cabling and installation in aerial, underground, and submarine configurations. Efficient and cost saving cable designs require optical waveguides to withstand tensile stresses over kilometers of length and years of operation. Since laboratory testing of kilometer lengths over periods of many years is untenable, the causes of mechanical failure and data analysis must be well understood for accurate extrapolations from laboratory test results to the long lengths and time spans characteristic of field deployment. The use of the Weibull probability function as a statistical tool for analysis of waveguide mechanical strength has prevailed. The Weibull function is usually employed to plot a cumulative failure probability versus stress at failure with dynamic loading, or time to failure with a static load. Data plotted accordingly is also representative of the sample gauge length of fiber tested and accurately predicts strength behavior for longer or shorter lengths. Fatiguing influence can also be extracted from the Weibull plots with additional statistical operations. Applications foreseen in the near future may very well call for the implementation of cable designs which must rely upon the optical fiber within to bear the burden of hundreds of thousands of pounds per square inch. To achieve such rigorous performance, manufacturing capabilities must be intelligently directed and to do so demands a sound comprehension of the principles of strength and fatigue in optical fibers.