Using a validated transmission model for the optimization of bundled fiber optic displacement sensors.

A variety of intensity-modulated optical displacement sensor architectures have been proposed for use in noncontacting sensing applications, with one of the most widely implemented architectures being the bundled displacement sensor. To the best of the authors' knowledge, the arrangement of measurement fibers in previously reported bundled displacement sensors has not been configured with the use of a validated optical transmission model. Such a model has utility in accurately describing the sensor's performance a priori and thereby guides the arrangement of the fibers within the bundle to meet application-specific performance needs. In this paper, a recently validated transmission model is used for these purposes, and an optimization approach that employs a genetic algorithm efficiently explores the design space of the proposed bundle sensor architecture. From the converged output of the optimization routine, a bundled displacement sensor configuration is designed and experimentally tested, offering linear performance with a sensitivity of -0.066 μm(-1) and displacement measurement error of 223 μm over the axial displacement range of 6-8 mm. It is shown that this optimization approach may be generalized to determine optimized bundle configurations that offer high-sensitivity performance, with an acceptable error level, over a variety of axial displacement ranges. This document has been approved by Los Alamos National Laboratory for unlimited public release (LA-UR 11-03413).