Electrical tree propagation: from stochastic models to a quantitative physical description

摘要

Considerable advances have been made in the understanding of electrical tree formation. In spite of the success of the stochastic models in predicting trees to exhibit a Weibull distribution of fractal structures, experimental evidence is in favour of a model based on deterministic chaos in a deterministic breakdown mechanism. Such an interpretation pulls together a number of disparate observations for electrical trees, namely, their decelerating growth rate, the sharp transition to runaway, the presence of strange attractors in their electrical discharge sequences, and their fractal nature itself. The D-A model fulfils the requirements for an electrical tree mechanism embracing deterministic chaos, and has also been shown to describe the observed field dependence of tree growth. Features of tree formation are given a physical interpretation, and are thus amenable to experimental evaluation. Transit time statistics are also predicted and the experimentally obtained distribution width is related to the contributory influence of electrical inhomogeneity.