<title>How to report light loss values for optical fibers used in fiber optic lighting applications</title>

Abstract

This manuscript attempts to explain why reporting a single value for light loss or a single wavelength-dependent attenuation graph is of minimum use to the lighting specifying community. Experimentally it was found that the light loss for different diameter fibers of the same type with the same illuminator varied from 1.1 to 1.5 percent per foot. This is mainly attributed to the higher number of reflections at the core-cladding interface when light travels through a smaller diameter fiber compared to a larger diameter fiber. In a similar experiment it was found that the light loss for the same fiber, size and material, on three different commercial illuminators with the same 150-watt metal halide lamp varied from 1.4 to 2.4 percent per foot. The numerical aperture through which light is coupled into a fiber and the spectral uniformity at the common-end of the fiber can affect fiber attenuation. It is shown here that the color separation of a metal halide lamp has minimal impact on the light attenuation and thus the greatest light loss effect is due to the launch angle. Reporting a family of attenuation plots, as function of fiber diameter and launch angle would be more useful for predicting light loss and also for developing high efficiency fiber optic lighting systems.