Search

The results of an investigation concerning a mechanism of brittle fracture in glass reinforced plastic (GRP) rods used in non-ceramic insulators (NCI) are presented. Commercial grade GRP rods and GRP rods from actual insulators were exposed to ultra-pure water (UPW) and acids while being subjected to mechanical stresses. The experimental results revealed that water has the potential of inducing stress corrosion cracking on the fibers and hence brittle fracture in the rods. It is observed that the fracture proceeded faster when the rods were exposed to UPW than when exposed to acids. Furthermore, a brittle fracture in an epoxy cross-arm, which was installed in a region where the formation of acids in the atmosphere can be neglected, is analyzed. Based on these evidences, it is postulated that the failure of in-service NCI in the brittle mode can occur under the influence of water and mechanical stresses, and that the failure is more likely to happen with water than with acids.

Glass Reinforced Plastic (GRP) rods used in Non-Ceramic Insulators (NCI) were examined for microscopic defects in order to evaluate their role in causing brittle fracture. The rods evaluated utilized different types of glass fibers and resin systems. Niemeyer's generalized approach to Partial Discharge (PD) modeling was used to investigate the development of electrical discharges inside the rods. The results indicated that the PD inception voltage was well above the highest system voltage used presently. Thus, it was concluded that the reactive species required for the formation of acids could not be developed inside the rod. The results also suggest that chemical reactions other than acid production can promote stress corrosion cracking of the rod and these may play a significant role in the failure process.

A new laboratory test for evaluating the tracking and erosion performance of HV outdoor polymeric insulating materials is described. The materials evaluated include various formulations of HTV (high temperature vulcanized) silicone rubber and polyolefin polymers. The test is based on combining some features of the ASTM D2132 DF (dust and fog) test and the ASTM D2303 IP (inclined plane) test. The new test employs IP test geometry, IP test equipment, and IP specimen plaques. The plaque is coated with a mixture of clay and salt similar to the contaminant of the DF test and identical to the contaminant used in the IEC clean fog test to rate ceramic insulators for use in contaminated environments. A liquid contaminant with a conductivity similar to that of the DF test is applied to the test specimen in the same way as in the IP test. Data collected from the field on the maximum concentration of insoluble and soluble ionic materials on surfaces of contaminated insulators which have been for 21 years in contaminated regions, provide a basis for choosing the minimum concentration of the solid contaminant to apply to the specimens. This choice also serves to define, at least tentatively, the geographical area where the results of this test have significance. It is expected that this test could be used also for screening materials and obtaining a relative ranking of the tracking and erosion resistance of various materials. Measurements of the leakage current via a computerized data acquisition system, and the discharge activity with a high speed camera were performed, and have resulted in a better understanding of the onset of material degradation.

In this paper, a methodology to build the required models for the performance comparison of materials that usually pass the standard inclined plane tracking and erosion tests, is presented. Samples made of silicone rubber and cycloaliphatic epoxy resin were evaluated to present the practical aspects of the proposed methodology. Factorial experiments, using the modified version of the inclined plane test method, were conducted to obtain the response variable. Generalized linear models provided the base for the proposed statistical models since the selected response variable was not normally distributed.