The strength degradation behavior of fused silica optical fiber is well known to be sensitive to the temperature and an apparent activation energy can be determined. In addition, it has been observed that the activation energy also depends on the applied stress and the nature of the environment. However, no consistent model for this behavior has emerged. We propose a chemical kinetics model which accounts for the temperature dependence of the dissociation of water which predicts that degradation should be faster in pH 7 buffer than in pure water. Static fatigue of fused silica fibers in both water and pH 7 buffer solution has been carefully studied as a function of temperature to test the model. The apparent activation energies are stress dependent, and, while the dependency is not clear, different environments give different dependencies. These observations support the proposed model.

The two general types of strength-reducing defects most commonly found in lightguide fibers are those due to chemical interactions of refractory particles with the hot glass and those due to mechanical damage resulting from indentation or scratching. In this paper we review some work on the mechanical damage in silica lightguide fibers and some of the approaches which have been taken to simplify the study of such damage. We are reminded that the indentation and scratching behavior of silica glass is not simple and that in all probability there are a variety of different types of flaws which may have different fatigue and aging behavior, and therefore result in different lifetimes. It is therefore suggested that either the actual proof test level flaws, or at least a wide variety of synthetic flaws, must be studied in order to develop sufficient background and experience to confidently predict fiber lifetimes.

Silica-clad optical fibers are proof tested in tension at levels of 700 MPa or greater to ensure product quality. The flaws passing proof testing generally are assumed to be ideal; however, some of these flaws are associated with refractory particles that are fully or partially submerged on the silica surface. Residual stresses are generated in the silica due to a thermal expansion mismatch between the particles and the host silica. These stresses can affect the crack growth characteristics and lifetime predictions for silica-clad optical fibers.

Residual stresses around refractory particles are quantified using analytical and finite element models. Zirconia particles generate the highest stresses at low aspect ratios. As aspect ratio increases, materials with higher elastic moduli, such as SiC, generate the highest stress levels. The residual stress fields are incorporated into crack stress intensity relations and their effect on optical fiber strength is evaluated for different refractory materials, particle sizes, and particle aspect ratios. The results are compared to the residual stress fields and stress intensity relations for model flaws created by indentation with diamond pyramids.

The analytic calculation and the measurement of thermal stresses have been performed for concentric three layered structures of a fiber-core, inner, and outer cladding. Especially the effects of F-doped inner cladding on the residual stress have been parametrically studied.

We examine the mechanical response of an optical fiber interconnect clamped (by epoxy bonding or soldering) at the ends and then subjected to a not-very-small lateral ends off-set. The term “not-very-small” is used here to indicate that the off-set is small enough to apply the elementary (linear) beam theory, but, at the same time, is large enough, so that the effect of the reactive tension has to be accounted for. The obtained results can assist a reliability engineer to make, if necessary, the appropriate changes in the photonic package design.

A solution to the nonlinear problem of bending of an optical fiber interconnect is considered for the case when the interconnect experiences large deflections due to a significant lateral ends off-set. Thissolution can be helpful in the analysis and design of interconnects subjected to a significant ends off-set. It can be used also to assess the accuracy of the predictions, based on the linear approach.

We examine a glass fiber “pigtail” in a gain equalization filter (GEF) splice box (SB) design. The fiber experiences bending on the surface of the SB cylinder. The objective of the analysis is to evaluate the “natural” configuration of the fiber, and to suggest, its optimized configuration for minimum curvatures and lower bending stresses. We show that the effect of such an optimization is the largest (about 20%) for cylindrical surfaces of large radii. The obtained solutions can be helpful in the analysis of optical glass fibers experiencing bending on cylindrical surfaces.

Many of the broken fibers in optical connectors, especially those that seem to occur over time without apparent provocation are found in the ferrule at the transition from entry cone to alignment capillary. This paper contends that many such breaks are due to local stresses caused by debris or some other relatively rigid imperfection in the transition neighborhood. Stress estimates from beam theory and from a finite-element model are presented along with some indirect experimental observations supporting the contention.

In this work we analyze the existing literature relating to the B-value and the inert strength measurements. The weak flaw statistic model for fiber lifetime estimation is considered with regard for the two-region crack growth.

The effect of the two-region crack growth model on the static and dynamic fatigue curves for strong and abraded fibers is studied. The additional requirements to static and dynamic fatigue experiments necessary for the correct determination of the fatigue parameters are discussed.

Optical fiber models for mechanical reliability require that the initial strength and crack growth parameters be measured. High speed testing allows one to investigate and model common high-speed processing events during fiber processing such as proof testing, coloring and cabling. In this study stressing rates ranging from 7×10−6 GPa/s to 1.5 TPa/s were accomplished using a universal testing machine, belt slide and pneumatic piston. When plotted in typical dynamic fatigue fashion the data shows curvature at the faster stressing rates. The presence of region II type crack growth is suggested as a possible explanation for this curvature. A multi-region crack growth model is used to extract crack growth parameters that are used to make comparisons with crack velocity results on bulk glass.

A review is presented of metallization processes of silica optical fiber. Fiber strengths and fatigue data are summarized and compared for liquid metal freezing, chemical vapor deposition, sputtering, electroless and electrolytic coatings. Both continuous in-line and termination-only processes are considered. Details of an electroless Ni-P plating process are presented which can be used to prepare fiber ends for soldering into hermetic packages.

Metal-coated fibers are irreplaceable in a multiplicity of uses owing to high hermeticity of the coating and its chemical and thermal stability. At present, the most widespread technique for the metal coating application onto a fiber is a so-called freezing technique, which consists in pulling a silica fiber through a layer of a metal's melt directly in the fiber drawing process. If the fiber temperature is lower than the metal melting point, a certain amount of the melt is frozen on the fiber surface in the form of coating. For this to occur, it is necessary that the duration of the fiber-melt contact did not exceed the time during which the fiber heats up to the metal's melting point; otherwise, the frozen metal will melt again. Because silica glass is poorly wetted with a majority of liquid metals, the coating in the latter case is likely to become discontinuous.

The mechanism of the coating of molten tin on a silica optical fiber is analyzed by a numerical method based on a three-dimensional conduction model.

We report, for the first time, optical fibers coated with Ni by the freezing technique. Ni-coated fibers demonstrate acceptable mechanical strength and effects related to magnetoactivity of the coating. Mechanical and optical characteristics of Ni-coated fibers along with their promising applications are discussed.

Hermetically metal coated fibers are candidates for applications at high temperatures. To date, various metals with melting point up to 1400 °C have been used to coat fibers by the freezing technique. However, the major problems consist of ensuring mechanical reliability and acceptable optical losses in metal coated fibers at high temperatures. We present an experimental study of the temperature effects at 20 – 1050 °C on the performance of optical fibers with copper coating. Specific fiber degradation mechanism have been revealed

While the telecommunications optical fiber technology is in its third decade of growth, and fibers that are aging-resistant and fatigue-resistant are highly desirable, basic physico-chemical aspects of fiber aging and mechanical strength degradation are not well understood. A key difficulty to understanding fiber strength degradation has been the unavailability of tools for direct assessments of the condition of fiber coatings on aged fibers. In this presentation, we report on our studies of the effects of DI-water aging on two experimental high n-value (fatigue-resistant) fibers and their coatings. We have found that the coatings of these high n value fibers are highly resistant to DI-water aging.

The properties of the optical fiber buffer coating play a critical role in the strength and handleability of optical fibers. Tests have been developed to measure a coating's resistance to mechanical damage. Two such tests are described here. The indentation rupture test involves indenting coated fiber with diamond pyramids and measuring the load at which the pyramid punctures the coating. Finite element analysis is used to determine the strain energy absorbed by plastic deformation of the coating. The toughness of the outer coating layer and the toughness of the inner coating/glass interface can be determined from the crack dimensions produced during rupture of the coating. Strength after mechanical abrasion is a procedure wherein a hollow tube of coating is removed from an optical fiber and subsequently bombarded with an abrasive grit. The effects of the abrasion event are measured by a tensile test of the tube. Results for both tests are obtained for two coating systems. The validity of the tests for measuring coating properties is discussed.

Reliability of lnGaAlAs mesa-waveguide photodiodes suitable for low-cost surface integration modules applied to access network systems is investigated. A passivation technique including silicone-based polymer encapsulation was used to ensure high reliability. Under a non-humidity condition of 200°C/10 V, a 9,600-hour stable operation was observed, and the estimated median lifetime is more than 100 years at the field service temperature of 85°C. A measured slight change in the dark current is explained in terms of hole accumulation at the semiconductor-insulator interface. Under a humid condition of 85°C/85% RH/10 V, there was no change in the dark current over 10,000 hours. The observed median lifetime at 135°C/85% RH/10 V of 1,500 hours would lead to an estimated median lifetime at 45°C/50% RH of about 50 years. These results suggest that the WG-PD is applicable to low-cost non-hermetic optical modules for optical access networks.

Many optical fiber sensor systems in outdoor applications are in operation and reliability aspects are a major concern. We installed multiplexed fiber optical Bragg grating sensor systems in a stay cable bridge and in a power dam both operational for about two years now. We report on design aspects of such systems which have several constraints: First, the sensing parts of a single multiplexed fiber had to be located at certain positions and in a specific orientation. Further, the sensors must be compensated for environmental influence, the fibers were not allowed to be bent too tightly, and the fibers had to be embedded and protected in a sensor body in a way to survive its installation in the monitored structure. Finally the sensor must be reliable to be a useful measurement device.

We present long term monitoring data and test results concerning the reliability of our systems and discuss benefits and disadvantages of the various applied sensor designs.