3:30pm - 3:50pmDielectrics II: 1
Electrostatic Surface Charge Decay of Floating Dielectrics
1Texas Tech University, United States of America; 2Pantex, Amarillo Tx. United States of America
Electrostatic surface charge accumulation on dielectric materials, followed by surface charge decay, is investigated. This work focuses on charging floating dielectric surfaces to the limit of electric breakdown in atmospheric air in humid and dry conditions, succeeded by the slow charge decay on the timescale of minutes to hours. The mechanisms leading to reducing the surface charge density include surface charge cancellation from ions attracted from the surrounding gas medium as well as charge migration along the dielectric. A 100 mm diameter sphere of varying materials (Teflon, Acrylic, and metal as a reference) was triboelectrically charged to tens of kilovolts and allowed to decay uninterrupted in relative humidities of 40% and 12%.
While the metal sphere charge decay was largely unaffected by the humidity, the dielectrics exhibited a much faster surface charge density decay in humid conditions, particularly when the surface charge initially covered only a fraction ~ 10 to 25% of the sphere. That is, the humid conditions cause moisture layers to form on the dielectric’s surface, impacting the surface conductivity of the material and allowing charge to redistribute along the surface. For instance, one finds an exponential dependence of the conductivity on the number of adsorbed water layers for Teflon and quartz reported in the literature.
For acrylic, in humid air, an initial drop in voltage upwards of 30% of the initial charging voltage was observed over 5 seconds. This rapid voltage drop is attributed to charge redistribution occurring on a much faster time scale than the air-ion to surface charge recombination. Further experimentation on decay behavior from uniformly charged dielectric spheres and partially charged spheres was conducted. This work provides details on the nonlinear surface conductivity, which is found to be electric field dependent.
3:50pm - 4:10pmDielectrics II: 2
Characterization and Modeling of electrostatic Discharges on floating dielectric Materials
1P3E Center, Texas Tech University; 2Mission Engineering Development Group, CNS Pantex; 3Facility Engineering Electromagnetics Group, CNS Pantex
Electrostatic Discharge (ESD) waveforms are measured for polymeric materials – PTFE, PMMA, PA6 – with high-speed (sub-nanosecond) current viewing resistors (CVR). These waveforms are used to create lumped element models which capture the behavior in addition to a comparison to a drift-diffusion numerical simulation. Cylindrical dielectric samples without a well-defined ground (i.e. samples are ‘floating’) are of particular interest in this study. The ~100 mm diameter samples are charged primarily via triboelectric means to high voltage – greater than 20 kV. The surface charge distribution is mapped before and after a discharge to determine energy lost to establishing the spark, conduction to ground, and radiation – captured with a B-Dot probe. A three-axis – R, Z, and rotational Phi – movement system was created to perform the mapping of charges and control the approach speed of the discharging electrode. Discharge electrodes consisting of spheres of 5-20 mm diameter and pointed electrodes with tip radii ranging from 0.1 to 1 mm are used with approach speed ranging from 10 to 150 mm/s. The radiated field from the B-dot probe exhibits a sharp peak at ~1.5 GHz, temporally coinciding with peak current. Discharge waveforms are similar in shape between materials and charging polarities; however, peak current and length change. For instance, PTFE charged negatively creates a spark 150-200 ns in length, whereas positively charged PMMA creates a spark 100-150 ns in duration. Peak currents, on average, are similar between materials and polarity, ~0.2 A, and peak dI/dt range from 0.3 to 0.57 A/ns for PTFE and PMMA, respectively.
Through pre- and post-mapping of the surface charge, the discharged area of the charged dielectric is captured. The discharged areas range from 4 to 8 cm2, with the charge maps revealing extended surface tracking mainly parallel to the symmetry axis of the cylinder.
With the spark length mainly dependent upon the approach speed for the same voltage, namely that faster approach speeds result in shorter sparks, one may expect lower speeds to exhibit lower peak current – that is, a longer spark would be suspect of losing more energy in establishing the spark. However, in experimental studies, this doesn’t always occur. While a Rompe & Weizel spark model with an RC object model captures the basic behavior of the discharges, the drift-diffusion model reveals the fundamental physics at play. The numerical simulation captures the electron and ion motion/generation/loss and includes modeling the emission process of charges from the dielectric surface.
4:10pm - 4:30pmDielectrics II: 3
A Finite Element Analysis Model for Internal Partial Discharges under DC Voltages
1Virginia Tech, United States of America; 2University of New Mexico, United States of America; 3Avalanche Energy Designs, United States of America; 4University of Manitoba, Canada
While the PD topic is booming for ac systems, it is immature for dc systems. Also, although much work has been done and significant progress has been made on PD measurement and detection techniques, this is not the case for PD modeling. To address these technical gaps, a finite element analysis (FEA) model for internal PD under dc is, for the first time, developed to help understand the mechanisms behind PD under dc voltages. The model is validated through experimental studies where testing samples, one with a void and another without any void, are built using 3D printing. The sample with a void has a void inside, which will cause internal PD. There is a good agreement between the measured and simulated magnitude of charges and the frequency of PD signals under dc voltages.
4:30pm - 4:50pmDielectrics II: 4
Federal University of Campina Grande, Brazil
The transmission of electrical energy in the world is predominantly provided by High Voltage Alternating Current (HVAC) lines. However, High Voltage Direct Current (HVDC) lines have gained space in power transmission over long distances. As the transmission lines, both in HVDC and HVAC, are susceptible to efforts from the environment and from the action of electromagnetic fields in transmission energy process. Such efforts wear out the materials that make up the system, especially the insulators, which can lead to failures and interruptions in the electricity supply. One of the main causes of interruptions, whether scheduled or unscheduled, is associated with the performance of electrical system insulation. In the case of electrical transmission and distribution systems, the insulation between the conductors and the towers is carried out by atmospheric air and by components called electrical insulators. Among electrical insulators, polymeric insulators have a prominent role due to their comparative advantages to ceramic insulators. To mitigate insulation failures that are related to polymeric insulators, it is recommended the use of predictive techniques to identify damages that could compromise the system. A prominent technique in the electrical system is infrared thermography. It suggests that the presence of defects in the polymeric insulators causes an increase in the local leakage current, which generates abnormal heating of the component. In AC, the technique is used by power utilities due to its electrical efficiency and usability in other equipment. In DC, operational experience using thermography to identify defective individuals has not been documented. This work presents a comparative analysis of the techniques for the application of infrared thermography in polymeric insulators in research on alternate (AC) and direct (DC) voltages. Therefore, electrical tests in AC and DC were carried out on polymeric insulators with different operating states. It was observed that polymeric insulators with degradations around the core, associated or not with fiberglass exposure, cause localized heating in the AC test. The generated heat identifies faulty regions in the insulators and allows effective monitoring of the evolution of damage to the polymeric coating. On the other hand, the results obtained from the insulators in DC did not identify a significant increase in temperature in any region of the insulator. The low value of the leakage current was the cause indicated for the thermal profiles obtained in the DC test. The results indicate that, in general, they use as monitoring techniques what is not possible to identify an isolated condition directly, except in cases where the defect is quite severe.
4:50pm - 5:10pmDielectrics II: 5
Effect of Titania Nanofiller on Electrical Tree of Silicone Gel
1Nagoya University, Japan; 2Toyohashi University of Technology
In recent years, to reduce the size and increase the power density of power modules, the inside of the module is exposed to high electrical fields. The high electric fields may cause electrical tree in the silicone gel used as an encapsulant inside the module. As a result, the electrical insulation strength and the insulation life of the silicone gel deteriorate. Therefore, it is very important to suppress the generation and development of electrical tree to extend the service life of the material. The addition of nanofillers to silicone gels has been found to be effective to suppress the generation and propagation of electrical tree. In a previous study, it was confirmed that the addition of titania had a significant effect on the suppression of electrical tree. However, the data showed some variability. Therefore, to increase the reliability of the verification of the suppression effect, we refined the evaluation conditions for the electrical tree. The breakdown experiments were conducted by setting several patterns of voltage increase speed, and it was confirmed that the higher the voltage speed increase, the higher the breakdown voltage. In addition, titania was added to the silicone gel to investigate the difference in the effect of improving dielectric strength depending on the amount of titania added. The samples were subjected to ultrasonic treatment to disperse the titania particles. As a result of the experiment, the sample with 0.1 vol% titania showed the highest breakdown voltage when the voltage was applied at 10 kVrms/sec. Further, we observed the progress of the electrical tree. The samples were centrifuged to remove micro-sized agglomerates. The filling rate after centrifugation was as small as 0.05 vol% or less, so it could not be measured, but since the sample itself was cloudy, it is considered that agglomerates or nanoparticles of a size that cannot be observed with an optical microscope remain. The AC voltage of 50 Hz was applied to the needle electrode at 0.14 kVrms/5 sec, 8.5 kVrms was continuously applied, and the progress of the electrical tree was observed using an optical microscope. As a result, the outline of the electrical tree observed in the nanocomposite gel was not significantly different from that of the unfilled silicone gel. We obtained, in addition, the electrical tree of neat gel stretched to 2 mm within 500 seconds. On the other hand, the electrical tree of NC gel stretched at the same speed as neat gel to 0.5 mm, and grew more slower than neat gel from 1 mm or more. Some trees did not extend to 2 mm. From these results, it is considered that the nanoparticles in the gel suppressed the progress of the electrical tree. In the future, we will continue our research on the mechanism of electrical tree propagation in silicone gel and nanofillers that can further suppress electrical tree.