We discuss a measurement method that aims to determine the radial distribution of the photoelastic constant C in an optical fiber. This method uses the measurement of the retardance profile of a transversely illuminated fiber as a function of applied tensile load and requires the computation of the inverse Abel transform of this retardance profile. We focus on the influence of the measurement error on the obtained values for C. The results suggest that C may not be constant across the fiber and that the mean absolute value of C is slightly larger for glass fibers than for bulk fused silica. This can, for example, influence the accuracy with which one is able to predict the response of optical fiber sensors used for measuring mechanical loads.

We measure the radial profile of the photoelastic coefficient C(r) in single-mode polymer optical fibers (POFs), and we determine the evolution of C(r) after annealing the fibers at temperatures from 40°C to 80°C. We demonstrate that C(r) in the fibers drawn from a preform without specific thermal pre-treatment changes and converges to values between 1.2 and 1.6×10^-12 Pa^-1 following annealing at 80°C. The annealed fibers display a smoothened radial profile of C(r) and a lowered residual birefringence. In contrast, the mean value of C(r) of the fiber drawn from a preform that has been pre-annealed remains constant after our annealing process and is significantly higher, i.e., 4×10^-12 Pa^-1. The annealing process decreases the residual birefringence to a lower extent as well. These measurements indicate the impact of annealing on the thermal stability of the photoelastic coefficient of POFs, which is an essential characteristic in view of developing POF-based thermo-mechanical sensors.

We determine the radial profile of the photoelastic constant C(r) in two single mode and one multimode polymer optical fibers (POFs), all fabricated from polymethylmethacrylate (PMMA). To determine C(r) we first determine the retardance of the laterally illuminated fiber submitted to a known tensile stress uniformly distributed over the fiber cross-section. Then we determine the inverse Abel transform of the measured retardance to finally obtain C(r). We compare two algorithms based on the Fourier theory to perform the inverse transform. We obtain disparate distributions of C(r) in the three fibers. The mean value of C(r) varies from-7.6×10 -14 to 5.4×10 -12 Pa -1 . This indicates that, in contrast to glass fibers, the radial profile of the photoelastic constant can considerable vary depending on the type and treatment of POFs, even when made from similar materials, and hence the photoelastic constant should be measured for each type of POF.

We discuss two algorithms to determine the value and the radial profile of the photoelastic coefficient C in glass and polymer optical fibers. We conclude that C is constant over the fiber cross-sections, with exception of silica glass fibers containing a fluorine-doped depressed cladding. In the undoped and Ge-doped parts of these silica glass fibers we find a consistent value for C that is slightly larger than in bulk silica. In the fluorine-doped trenches of the absolute value of C decreases with about 27%. In polymethyl methacrylate optical fibers, the value of C significantly varies from fiber to fiber.

No abstract