Process and Health Monitoring of FRP by Rayleigh-Scattering Based Distribution Optical Fiber Sensors

Distribution optical fiber sensors are attractive because of capability of distribution sensing of strain for large-scale FRP structures. Recently, Rayleigh-scattering based distribution optical fiber sensors, which has high spatial and strain resolution, have been developed. Therefore, it is expected that local strain distribution produced by such as local damages or unevenness of temperature distribution can be monitored by the Rayleigh-scattering based sensors. In the present study, feasibility studies of process and health monitoring of FRP by the Rayleigh-scattering based sensors were conducted. In the monitoring of VaRTM process of textile GFRP, flow front of resin impregnation and process-induced strain were monitored. From the monitoring results of impregnation process, it was found that the flow-front of the resin could be observed as the maximum compressive strain in the strain distribution. From the strain curves after start of heating process (40 minutes), the optical fiber sensor could measure thermal strain by heating from 40 to 100 minutes, curing shrinkage from 100 to 150 minutes and thermal shrinkage by cooling from 150 to 170 minutes successfully. Therefore, it can be concluded that the Rayleigh-scattering based sensors could monitor impregnation status of resin and distribution of process-induced strain during VaRTM process. In the health monitoring, damage identification of cross-ply GFRP laminates with delamination by the attached distribution optical fiber sensors was carried out. Three types of specimens, no delamination, delamination between the central layers and delamination between the outer layers were prepared. The distribution sensor was attached on the bottom surface. From the results, it appeared that the strain varied largely on the delaminated section and the measured distribution is similar to the analyzed one. Therefore, it was found that the Rayleigh-scattering based sensors could identify delamination from surface strain distribution.