An application of a plastic optical fiber sensor and genetic algorithm for structural health monitoring

Abstract

This article demonstrates the potential of a new extrinsic plastic optical fiber (POF) sensor for vibration-based structural health monitoring (SHM) applications. A simple genetic algorithm (GA) is used in conjunction with the POF sensor. The construction and principle of operation of the intensity-based POF sensor used in the study is described, highlighting the advantages of the present sensor over previous designs based on glass fiber. A series of mechanical tests are conducted to evaluate the performance of the POF sensor and the results demonstrate the potential of the sensor for SHM. To assess the potential of the POF sensor for vibration-based damage identification, the sensor is attached to a beam to measure the global dynamic response of a carbon fiber composite beam. The POF sensor is surface-bonded to the cantilever beam and collocated with a polyvinyl idine fluoride-based piezofilm sensor for comparison. The time-history responses of the sensors following an impulse-type excitation of the beam are obtained and analyzed using classical Fourier transform techniques to identify the modal frequencies of the beam. Based on the basic principle that changes in structural properties will lead to changes in its vibration signatures, the sensor is used to monitor and acquire the vibration signatures. To simulate the vibration signature shifts, weights are used which are located sequentially at specific distances along the beam. The global dynamic responses of the beam are recorded using an oscilloscope following the excitation of the beam. The results show that the POF sensor is capable of detecting the relevant modal frequencies. Using a GA approach, the mass and location of the weights are identified based on the optical signal acquired through the POF sensor. This article demonstrates the potential of the POF-GA system for damage identification in a simple cantilever beam. © 2006 SAGE Publications.