Classification of Aged Epoxy Micro–Nanocomposites Through PCA- and ANN-Adopted LIBS Analysis

Abstract

Epoxy micro–nanocomposite specimens are subjected to different levels of gamma irradiation and water aging to analyze their degradation behavior after aging. The plasma temperature calculated from the laser-induced breakdown spectroscopy (LIBS) spectral data tends to decrease, and the crater depth after laser ablation tends to increase with increment in the level of aging, reflecting the reduction in surface hardness of the epoxy specimens after aging. Water aging creates a more severe impact on the surface hardness of the specimens, compared to gamma irradiation, in the range studied. The normalized intensity ratio of H(I)/Si(II) is in direct correlation with the water diffusion coefficient. Machine learning techniques, such as the principal component analysis (PCA) and the artificial neural network (ANN) analysis, have been performed on LIBS spectral data in order to classify the unaged, gamma-irradiated, and water-aged specimens. The number of neurons in the hidden layer of the neural network architecture has been optimized based on classification accuracy and the number of epochs required to converge. The ANN-adopted LIBS analysis is successful with good classification accuracy compared to PCA, for categorizing the level and type of aging in epoxy micro–nanocomposites as observed in this work.