Design of an Embedded Sensor Network for Manufacturing Process Monitoring, Quality Control Management and Structural Health Assessment of Advanced Composite Structures

Abstract

This paper proposes the use of an embedded network of fiber optic sensors for process and Structural Health Monitoring (SHM) of Resin Transfer Molded (RTM) composite structures. A single sensor network is used at each stage of life of a RTM composite panel: flow monitoring, cure monitoring and health monitoring. A laboratory scale RTM apparatus was designed and built with the capability of visually monitoring the resin filling process. A technique for embedding fiber optic sensors into the mold has also been developed. Both Fiber Bragg Gratings (FBG) and Etched Fiber Sensors (EFS) have been embedded in composite panels using the apparatus. Etched Fiber Sensors have the capability of detecting the presence of resin. The sensors have proven to be capable of detecting the presence of resin at various locations as it is injected into the mold and have the capability of being multiplexed with FBGs thus reducing the number of ingress/egress locations required per sensor. Two FBGs and three EFSs were embedded on a single optical fiber. Tensile test specimens that contain embedded FBG sensors have also been produced with this apparatus. These specimen and embedded sensors have been characterized using a strain gage and a material testing machine. FBG sensors have been embedded into composite panels also in a manner that is conducive to detecting Lamb waves generated with a centrally located PZT. To detect Lamb waves a high speed, high precision sensing technique is required for embedded FBGs, since these guided waves travel through the material at very high velocities, presenting relatively small strain amplitudes. A technique based in a filter consisting of a second FBG was developed. Since this filter is not dependant on moving parts, it does not limit the velocity or frequency at which the tests can be performed. Preliminary tests performed using this filter showed that it is possible to detect Lamb waves with amplitudes smaller than 1 microstrain. A damage detection algorithm has been developed and is applied to this system in an attempt to detect and localize damages (cracks and delaminations) in the composite structure.