This work presents a new method for monitoring the preload torque in a composite bolted connection using an embedded fiber Bragg grating sensor. A unique washer was designed to impose a specified nonuniform strain field across the grating, causing distortion in the reflected optical spectrum. Using the full-width at half maximum bandwidth of the Bragg reflection spectrum as the preload-sensitive feature, it is shown that this feature increases monotonically-and quite linearly-with increasing applied bolt torque. It is also demonstrated that, although distorted, the spectral structure of the sensor is maintained such that it is simultaneously able to function in its typical use as a uniaxial strain sensor, thus essentially creating a dual-purpose sensor. The computational design approach is validated with a prototype experiment.

In recent years, the use of composite materials has helped achieve ever-increasing performance requirements in marine, aerospace, and civil structures. A parallel interest in the structural health monitoring (SHM) of composites has developed to further improve performance by reducing overall lifecycle costs. In this work, a network of embedded fiber Bragg gratings (FBG) is employed as part of a damage detection system for an impact damage scenario in a composite laminate material system. Delamination damage is incrementally introduced into the laminate via repeated impacts with a drop weight pendulum system. Using vibration time histories between impacts from a simulated, pseudorandom operational loading, damage sensitive features were extracted and placed within a Mahalanobis distancebased discrimination framework. The statistical modeling for hypothesis testing is also presented to give a full, systems-level approach to a damage detection system from data acquisition to ultimate decision making.