Masters Theses
Date of Award
8-1992
Degree Type
Thesis
Degree Name
Master of Science in Landscape Architecture
Major
Electrical Engineering
Major Professor
Marshall Pace
Committee Members
T. V. Blalock, Igor Alexeff
Abstract
A diagnostic consisting of specialized experimental techniques as well as a sensitive detection system was developed that could provide a real-time, non invasive record of water tree initiation, growth, and transition to an electrical tree. The experimental techniques included phase resolution, bandwidth and time constant adjustment, and several digital signal processing (DSP) techniques. The principle is to detect image currents in two electrodes, via two very low noise, wide bandwidth detectors, using the normal configuration of electrically stressing a dielectric between two electrodes. The two electrodes monitor a polyethylene sample in which water treeing is occurring, one electrode monitoring an "active" needle printed treeing region and the other monitoring a normal, or "reference", region which has essentially no water tree activity. Calibration showed that a signal originating in the sample at one probe was detected by that probe exclusively, but signals originating elsewhere were detected simultaneously by both. This experiment yielded two distinct electrical signals that were shown to originate only in the water treeing region. The signals detected were fast and slow current pulses,each exhibiting distinct identifying characteristics. Fast current pulses typically transferred 1.2-2.0 fC of charge and had widths of 25-30 nsec, while slow pulses typically had amplitudes of up to 125 nA with decay times as long as 55 μsec.
Recommended Citation
Dorris, David Leon, "Detection of electrical pulses during water tree growth in polyethylene. " Master's Thesis, University of Tennessee, 1992.
https://trace.tennessee.edu/utk_gradthes/12106