Early fiber reliability efforts concentrated on the development of manufacturing processes for the production of long lengths of strong fibers. More recently new devices and packages have been developed. Their successful deployment requires not only engineering procedures for the production of strong fibers, but also a much better understanding of the behavior of these fibers in a variety of situations.

The strength measurements of optical fibers is performed usually in tensile and in two-point bending. In tensile the weak flaw that will fracture the fiber has its effective length perpendicular to the length of the fiber. In two-point bending the maximum stress in the fiber will fracture the worst flaw which effective length is perpendicular to the length of the fiber. But the optical fiber manufacture process occurs in the length line and most defects originated during the process is aligned in length, and is not fully tested using two-point bending or tensile. Another important practical aspect not investigated is related with the low torsion stress present during long time in installed cables, that can cause fracture by fatigue. In torsion the maximum tensile stress is on planes forty-five degrees oriented with respect to the length of the fiber and it can better detect those defects. In this study dynamic fatigue torsion tests were performed and compared with other mechanical tests. The results show a good agreement between fracture distributions for fibers under torsion and under two-point bending.

Comparison of high-speed strength data for weak (abraded, contaminated and indented) and pristine fibers was performed. It was shown that fatigue behavior of abraded fiber practically coincides with that of the fiber contaminated by zirconia powder and is close to that of indented fiber. The fatigue parameters obtained for strong pristine fiber cannot be used to obtain the correct prediction of fiber strength after proof testing. A two-region power law model was used for mathematical description of these results and the fatigue parameters for three types of weak fibers were obtained.

Ga-La-S glasses have been selected as important candidate materials for fabrication of infrared optical fibres. A study has been conducted to assess possible influence of the thermo-optical properties of Ga-La-S glasses on the ability to produce strong optical fibres. The work was based around Ga-La-S compositions containing small additions of oxide producing 0.65 wt% [O] in the glass. These low oxide Ga-La-S glasses have previously proved difficult to draw into strong optical fibres. The visible absorption edge of the Ga-La-S glasses was measured and found to shift to longer wavelengths as the temperature increased, moving from 550nm at room temperature to 700nm at 600oC. At fibre drawing temperatures, around 660oC, the high frequency absorption edge had shifted from visible to infrared wavelengths significantly affecting the ability to draw strong fibres using conventional, radiative heating. Rods, with a 5mm diameter, of the Ga-La-S glasses were fabricated via an extrusion method producing a high quality surface finish suitable for fibre drawing. A conventional fibre furnace employing radiative resistance heating and a novel convective fibre furnace were both used to drawn fibres from the Ga-La-S performs. Fibres drawn using the convective furnace showed an improved surface quality and strength compared to fibres drawn using the radiative furnace. Initial strength measurements of uncoated fibres, drawn using the convective furnace, tested under ambient conditions show strength in the order of 0.7GPa which are comparable to the strengths for coated fibres of other chalcogenide glasses tested under liquid nitrogen.

A new type of optical fiber has been developed with all pure silica in both core and cladding. The cladding is a nano porous silica produced on line from an oligimeric organo-silicate by a modified sol-gel technology. Characteristics, mainly mechanical properties, are described. The strength and fatigue of these optical fibers are very good, even without additional protective jackets. The nano porous silica is also being evaluated as an outer coating on all silica optical fibers. Unjacketed fibers have mean Weibull strengths in bending of 6.5 to 7.6 GPa with Weibull slopes in the 40 to 60 range. Strength decrease with decreasing strain rate is similar for both jacketed and unjacketed fibers. Static fatigue results using mandrel wrap tests are also presented. Dynamic and static fatigue parameters appear to be essentially the same with values around 20. A thin polyimide jacket does improve some of the mechanical properties. Results for nano porous silica ‘buffer’ over a silica/fluorosilica core clad structure are also presented. Effects of water and dry environments are presented, including results of short to intermediate term aging in boiling water. Possible mechanisms to explain the strength and fatigue behavior are discussed in light of these fibers’ unique structure.

Optical fiber may experience cyclic stresses at frequencies ranging from a few hertz in aerial cables to over a kilohertz due to vibration of machinery. The fatigue behavior of brittle materials typically gives times to failure that correspond to a suitably time-averaged applied stress and is independent of the frequency. Previous studies have been limited in the frequencies used but generally show agreement with this simple model. In this paper we describe results for the cyclic fatigue behavior of high strength fused silica optical fibers as a function of stress amplitude and frequency in the range of zero to 100 Hz. The results confirm that fatigue of this material is indeed accurately described by the subcritical crack growth model and the results are shown to be frequency independent in the range studied.

The reliability of the network optical fibers is a critical issue for telecommunications. New investigations methods have been developed within a cooperative program supported by France Telecom. They include low coherence interferometry and optical near field microscopy in association with classical analysis tools such as Scanning Electron Microscopy and Shear Force Microscopy. One aim of the study is the localization and the characterization of the defects from which failure originates. The aging effect has been investigated in silica fibers immersed in desionized water at 65 °C and 85 °C for different times : from 3 to 12 months

In the Power Law Crack Growth Theory the strength reliability of optical fibers is based on some mechanical fiber parameters. The n parameter, or strength susceptibility parameter, is the most important because it is a two-order exponent number used to calculate the survival lifetime. To measure this parameter different mechanical tests can be performed, among them the dynamic fatigue in tensile and the static fatigue in two-point bending. Static fatigue in the two-point bending technique using glass tubes to accommodate the samples does not require a lot of space and it can easily be set in different harsh environments. But still some problems can affect the results of this technique: the internal diameter irregularities of the tubes, the fiber damage introduced to the samples in the lower diameter tubes, and the uncontrolled stress rate during the loading process for all tube diameters. Using a two-point bending apparatus, and faceplate supports, precise static fatigue bending tests were performed measuring the n error calculated with the tubes. This apparatus allowed us to hit the maximum stress under very well controlled stress rate and no sample manipulation during the entire process.

The lifetime prediction model that has been developed for brittle materials implicitly assumes an initial flaw population from which, under the influence of an applied or residual stress, cracks grow to failure. The relationship between crack growth rate and stress intensity factor is assumed to be a power law expression. While the tests used to predict lifetimes, i.e., Weibull distribution, inert strength, and dynamic fatigue, can all be applied to semiconductors without apparent discrepancies, analysis of the crack growth behavior in Si, GaAs, and InP shows that assumptions in the lifetime model are violated for these materials. Consequently, lifetime predictions based upon this model will be wrong.

In this paper we present results obtained from computational studies on a set of experimental data concerning the strength of commercially available silica optical fibres. We report on the distribution of estimated values of the Weibull modulus using a variety of Weibull estimators and sample sizes. We compare the distribution of the estimated Weibull parameters from experimental data to analogous distributions obtained using Monte Carlo methods. We show that for small sample sizes <40 the maximum likelihood method produces a skewed distribution of estimated Weibull parameters. There is a small but finite probability that the Weibull modulus will be overestimated by a factor of 2. We also show that the experimental data are in reasonable agreement with our Monte Carlo simulation for linear regression methods. However, for maximum likelihood methods the peak of the distribution of estimated Weibull modulus can be seen to shift for the strength data but the peak is more stable for the Monte Carlo data The increased bias of the distribution of estimated Weibull modulus from experimental data must be considered when inferring the reliability of optical fibers from strength data.

Weibull analysis of tensile strength data is routinely performed to determine the quality of optical fiber. A typical Weibull analysis includes setting up an experiment, testing the samples, plotting and interpreting the data, and performing a statistical analysis. One typical plot that is often included in the analysis is the Weibull probability plot in which the data are plotted as points on a special type of graph paper known as Weibull probability paper. If the data belong to a Weibull probability density function, they will fall approximately on a straight line. A search of the literature reveals that many Weibull analyses have been performed on optical fiber, but the associated Weibull probability plots have been drawn incorrectly. In some instances the plots have been shown with the ordinate (Probability) starting from 0% and ending at 100%. This has no physical meaning because the Weibull probability density function is a continuous distribution and is inherently not bounded. This paper will discuss the Weibull probability density function, the proper construction of Weibull probability graph paper, and interpretation of data through analysis of the associated probability plot.

The fiber to the home technology is already in use and more fibers are necessary to carry all the demanded information. Ribbon optical cable, present in the market since the seventy’s, is the right technology to support those needs. Ribboned fiber mechanical strength is currently characterized using tensile tests, where the required handling affects the results. In this study, we demonstrate the use of two-point bending tests to characterize ribbon strength. By this method, handling was minimized and samples aged in harsh environments could be tested. Normal and Weibull Probability Distributions were used to characterize and compare the strength of two ribbons, and of the fibers used to construct them. Both distributions were efficient in describing the strength during an aging process done in harsh environment. Keywords: Ribbon Cable, Two-Point Bending, Dynamic Fatigue, Weibull, Normal.

Polymeric UV curable coatings have been successfully employed over 25 years to protect freshly drawn optical fibers from mechanical damage and to prevent microbending losses. Although present dual acrylate coatings provide satisfactory protection, they are water permeable and do not effectively protect the fiber surface from water corrosion. Rapidly growing market calls for next generation protective materials that would provide with enhanced fiber reliability even in hot, humid and other harsh environments. The general concept of such “hermetic” polymeric coatings design is presented. The key aspects of the design include the use of interpenetrating polymeric network (IPN) and organic-inorganic silica based hybrids derived by sol-gel process. Such coatings may offer superior moisture barriers and surface passivation while remaining otherwise equivalent to commercial acrylates from a processing perspective. Simultaneously, these coatings may offer other potential advantages such as significant thickness reduction, higher temperature performance and lower thermal expansion coefficients.

It is now well known that fused silica optical fiber can suffer from enhanced strength degrathtion after prolonged exposure to aggressive environments. This is caused by corrosion of the glass surface by moisture leading to roughening, strength loss, and, potentially, problems with handleability. It has been found that addition of nanosized silica particles to the polymer coating can improve the long term mechanical reliability by slowing corrosion and delaying the onset of strength loss. However, previous studies have shown that addition of these particles can lead to unacceptably high added optical loss, when measured using the "basketweave" test. In this work, it is shown that the added loss caused by coating additives can be reduced by improving the mixing and dispersion ofthe silica powders in the polymer. It is further shown that well dispersed powders still substantially improve the long term fatigue and aging behavior. This clearly shows that coating additives can improve the mechanical reliability without significantly degrading the optical performance.

Carbon coatings have been applied to silica lightguide fibers for many years in order to provide hermeticity to both water and H2. While there was a flurry of activity in this technology in the late 80’s and early 90’s, it appears that these coatings are little used today. The reasons for this are the increased cost associated with their use, the perception that they are unnecessary, and perhaps most important, the lack of complete understanding and control of their properties. In this paper we will review the history, preparation, structure and properties of these coatings in an effort to clarify some of the issues surrounding their production and use.

In earlier work, diffusion of moisture through polymer coatings was modeled by using an analytical solution to the diffusion equation and so was only applicable to the simplest cases, e.g. cylindrically symmetric Fickian diffusion. In this work the limitation of the analytical approach is avoided by the use of finite element analysis. However, finite element programs do not usually implement matter diffusion, and therefore it has been modeled by analogy with thermal conduction.

In this paper we present tensile and 2-point bending strength results on 5 cm. (2-inch) sections of mechanically stripped fiber. We show that the ‘weakest-link’ model is apparently not obeyed since the low strength mode of the resulting distributions have different widths, i.e., the Weibull m-values for the tensile and 2-point tests are ~2 and ~7, respectively. This simply means that the two measurements are sampling different flaw populations. The importance of understanding this behavior and applying that understanding in practice is emphasized.

The Strip Force test is widely used in the fiber optic telecommunications industry. This force is connected with the mechanical properties of the coatings, and in some extent with the adhesion between the coating and the glass. If the polymer coating fails by cracks or delamination with water penetration, it can cause fiber strength degradation, because of the glass humidity susceptibility. Until now there has been no correlation found between strength degradation, strip force and coating damages. In this study the strip force and the strength were measured for different fibers aged in harsh environment for long periods of time. Short and long samples were used. Short samples are like coating damaged fibers opened to the humid environment, and the long samples simulated long length of fibers inside cable structures. The results of the tests point out the everlasting life of the strip force and coating quality, opposing to the strength degradation of fibers under long term aging.

We are in the process of developing examples of acceptance testing design, and data analysis for connectors, for the IEC standard for reliability in passive optical devices. In this paper we describe the progress we are making in combining existing evidence, generating mechanistic hypothesis, and developing from these hypothesis valid experiments to use as acceptance tests for connectors. The connectors we are considering use epoxies or other adhesive polymers to bond the fiber into a ferrule, typically ceramic, at the interconnection. We show how the demarcation approximation, which allows us to approximately map out the physical processes that can occur for a given set of exposure conditions and time, is vital to the development of these kinds of tests.

We developed a high-strength proof-test for critical sensor applications, for which the reliability of fibre Bragg gratings (FBGs) strain sensors had to be guaranteed with a survival probability of more than 99.9 % over 5 years. Combined with statistical acceptance tests, in which both producer and user of FBGs agree, it will help to increase the quality and reliability of FBG sensors.

The tensile strength of fiber Bragg gratings is dependent on the type of UV laser exposure. The basic conclusion for the traditional method of producing gratings (exposure in the near-field region of a phase mask) is that the pulsed KrF excimer laser (248 nm) damages the fiber and the continuous wave frequency-doubled argon ion laser (244 nm) does not, provided that the fibers are handled carefully. Using the excimer laser at a low fluence (~5 mJ/cm2 pulse) and hydrogen loaded fiber, we demonstrate that Bragg gratings with an index change of 1.25x10-4 can be written. Although this index change is not enough to write a highly reflecting WDM grating, it is enough to write a weakly reflecting pump stabilization grating. The tensile strength of these fibers follow a Weibull distribution similar to pristine fiber with a median tensile strength of ~4.4 GPa (640 kpsi). A small percentage of the fibers are minimally damaged. As the fluence is increased, the median tensile strength decreases and the variability increases. The probability of damage from the laser as a function of the laser intensity suggests a damage mechanism related to laser-induced dielectric breakdown.

The strength reliability of spliced optical fibers is normally characterized using the tensile tests. When the fiber is under two- point bending, the glass surface area under maximum stress is very small and sharply localized, and it would be necessary exactly to situate the splice in that area, because that two-point bending is not used to study the strength of the splices. But an important question is related with the strength of spliced fibers: the strength degradation of the extremities of the fibers stripped during the splice operation. In those small lengths of fiber, around four centimeters, it is possible to use the two- point bending technique. In this work the two-point bending technique was used to measure the extremities strength of spliced fibers and a study of the flaw population presented in that vicinity is made comparing with tensile tests. The splices were prepared using an automatic method to strip, cleave and clean the fiber extremities.

Optical fibers are being widely studied for use as embedded sensors in composite materials. Appropriately chosen sensors can monitor parameters such as strain and temperature, either during curing or over the product life cycle for continuous health monitoring. In this work, the host material is a syntactic foam that polymerizes at high temperatures, rather than a laminated composite material. The objective of this work is to report on the mechanical strength of an optical fiber after exposure to the temperatures encountered during syntactic foam production, and after exposure to the precursors of the syntactic foam. These parameters are expected to affect the mechanical reliability of the sensor after foam polymerization is complete, during the articles’ life cycle. Results show two regions on a Weibull plot, possibly due to the testing procedure. Importantly, the fiber strength is higher than the anticipated load to failure for a fiber embedded in a syntactic foam, as required to achieve the desired failure mechanism.

In recent years optical fibers have been developed increasingly for sensor applications. In many of these applications optical fiber Bragg gratings are embedded in a matrix material as temperature or strain gauges. In effect, embedded optical fiber sensors are single-fiber composite (SFC) materials. The mechanical properties of fiber-reinforced composite materials are an important area of research, as well as the mechanical reliability of fibers used for sensors. The multiple fragmentation, with increasing load, of the fibers in SFCs is controlled by transfer of stress from the matrix to the fiber. The number of fiber fragments increases rapidly with increasing strain after the initial break with the fragment length decreasing until a limiting fragment length is reached. Analysis of the fragmentation process allows information to be obtained on the strength of the optical fiber on different length scales and on the transfer of stress from the matrix to the fiber. In this paper we present the issues surrounding mechanical tests on an optical flber/matrix SFC and the mechanical reliability of fibers for sensors. We consider the experimental difficulties and issues which must be addressed for a full understanding of interface effects and their effects on mechanical reliability.