Assessment of Elongation at Break of Silicone Rubber Insulation by NIR Spectroscopy

The power system has a peak demand for power cables because of the acceleration of urbanization in recent years. However, the huge number of cable and cable accessory suppliers results in highly variable product quality, which threatens the stability of the power system seriously. Therefore, it is necessary to test on the insulation of power cables and cable accessories. The traditional testing methods, however, are costly and time-consuming. Moreover, some testing methods that rely on slice sampling testing may damage the insulation of cables and their accessories, failing to meet the demand for non-destructive testing of power cables. Nearinfrared (NIR) spectroscopy, as a non-destructive and quickaccess method, can achieve this target easily. In this paper, sheet-shaped silicone rubber samples' NIR spectra and elongation at break were acquired. Partial least squares regression (PLSR) was employed to establish correlations between the NIR spectra data of silicone rubber and its elongation at break. In addition, wavelength selection methods were used to select the wavelengths which have the highest correlation with the target properties. Finally, the near infrared spectrum data of four silicone rubber samples were input into the prediction model to prove the accuracy of the model. It is found that the prediction error is less than four percent. Thus, through this process we can predict the elongation at break of silicone rubber by using near infrared spectra.