Characterization of tree growth in filled epoxy resin: the effect of filler and moisture contents

The effect of silica filler and moisture absorption on the structural characteristics of tree growth in epoxy resin has been investigated. The physical properties of the resin were varied by changing its filler concentration and immersed in distilled water at 50degC. The fractal dimension of the electrical tree and its relationship with filler and moisture contents were determined. The damaged area of tree in various concentrations of filler was also estimated. Furthermore, the lacunarity of tree with identical fractal dimension was calculated as a complement of fractal dimension characterization. In addition, moisture absorption and diffusion properties of the specimens used were also examined. It is considered that the filler would create such an obstruction to the tree growth both in moisture and non-moisture conditions. Under non-moisturized condition, the more filler concentration, the more obstruction would be generated, leading to the significant suppression of tree growth. Likewise, the introduction of filler brought a rise in fractal dimension due to the increase of branches. It is concluded that the existence of filler makes the tree structure more complicated by introducing obstacles to tree propagation, leading to the high fractal dimension of the tree. In addition, it was found that the fractal dimension of the tree was related to the fractal dimension of the composite material including filler particles. On the other hand, the lacunarity of "without-filler" specimen would be lower than that of in filled specimens. It is considered that the tree growth in filled specimen would have "localized-branching" at the interface between filler and resin leading to the heterogeneous inherent structures of the trees. As a complement of the complexity characterization by fractal dimension, the lacunarity can be a useful tool to quantify the heterogeneous structures of trees growth in composite materials. The moisture could hardly be absorbed into the filler compare- - d to the resin. It is shown by the higher value of diffusion coefficient of without-filler specimens than with-filler. 20 h of immersing is appropriate to achieve 40% in moisture content at the needle tip