The application of near infrared spectroscopy in the quality control analysis of glass/phenolic resin prepreg

During the manufacture of glass/phenolic resin prepreg cloth, the feasibility of near infrared (NIR) spectroscopy as a technique for the quality control analysis of the resin content, the volatile content and the resin pre-curing degree has been verified. The partial least square (PLS) regression was used to develop the calibration models by utilizing several different spectral pretreatments. The optimum models had determination coefficients (R ²) of 98.29 for the resin content, of 99.50 for the volatile content and of 97.66 for the pre-curing degree, respectively. The root mean square errors of prediction (RMSEP) for the resin content, the volatile content and the pre-curing degree were 0.376%, 0.169% and 0.105%, respectively. The results of the paired t-test revealed that there was no significant difference between the NIR method and the standard method. In the manufacture process of the prepreg cloth, the NIR on-line monitoring results were used to be the instructions for the quality control.