Application of PCA for quantitative the resin and volatile content of glass-fiber-reinforced phenolic resin prepreg by near-infrared spectroscopy

The near-infrared reflectance spectroscopy of glass-reinforced phenolic resin prepreg has been studied. The curing process of phenolic resin was investigated with principal component analysis (PCA) method. The mathematical models between near-infrared spectral parameters and quality standard of glass-fiber-reinforced phenolic resin prepreg (resin content and volatile content) are built up and the effects of the analysis accuracy are investigated by using the principal component regression. It was found that root mean square error of prediction for the resin content and volatile content is 0.6114 and 0.1885 respectively. The results show the feasibility of the near-infrared analytic technology for detecting the quality of composite prepreg.