Effect of Thermal Aging on the Dielectric Strength of Isoflex
Abstract Generated abstract
The study examines how thermal aging affects the dielectric strength of Isoflex, a composite insulation material based on polyimide film, glass fabric, and polyether epoxy varnish. Breakdown tests under industrial frequency alternating voltage were performed on Isoflex specimens aged at 200, 250, and 300°C, with additional tests on polyimide film and measurements of partial discharge characteristics. The electrical strength remained essentially unchanged after 1000 hours at 200°C, but at 250 and 300°C it exhibited a time dependent minimum that shifted to shorter aging times at higher temperature. The results suggest that this minimum is caused by thermal destruction of the binder, gas cavity formation, delamination, and associated partial discharges, followed by some field equalization as conductive degradation products reduce surface resistance.
Full Text
EFFECT OF THERMAL AGING ON THE ELECTRICAL STRENGTH OF ISOFLEX
V. I. Merkulov, T. N. Baskakova
(Presented by the scientific seminar of the Department of Electrical Insulation and Cable Engineering)
In recent years, composite materials based on polyimide and polyethylene terephthalate films have found wide application in electrical machine building as electrical insulation [1]. Among domestic composite materials, isoflex is of particular importance as a material possessing high heat resistance and high mechanical and electrical strength. The high physicochemical, mechanical, and dielectric properties of this material make it possible to use it as slot insulation in electrical machines and make it possible to automate the technological operations involved in winding the stator.
Taking into account that the slot insulation of electrical machines is subjected to mechanical and thermal effects, we investigated the electrical strength of isoflex during thermal aging.
The electrical strength of isoflex was studied at \(T = 20^\circ\mathrm{C}\) under alternating voltage of industrial frequency in a system of electrodes 20 mm in diameter.
Fig. 1. Dependence of \(E_{\mathrm{br}}\) of isoflex on the time of thermal aging \(\tau\) at different aging temperatures: 1 — \(200^\circ\mathrm{C}\), 2 — \(250^\circ\mathrm{C}\), 3 — \(300^\circ\mathrm{C}\).
As specimens for the study, isoflex-3N plates measuring \(100 \times 100\ \mathrm{mm}^2\) were taken, consisting of two layers of polyimide film—PM, 40 μm thick—and glass fabric impregnated with polyester-
with EP-934 epoxy varnish. To prevent flashover of the specimens, the electrode device proposed by Rizhinashvili [2] was used. The thickness of the specimens was measured on an IZV-2 vertical optimeter with an accuracy of ±0.5 μm.
The results of tests of the electrical strength of Isoflex during thermal aging are presented in Fig. 1. Each point corresponds to 30 specimens.
It is seen from Fig. 1 that, at an aging temperature of 200°C, the electrical strength of Isoflex is practically independent of aging time over 1000 h. This is confirmed by statistical verification of the test results using the Wilcoxon test [3]. At aging temperatures of 250 and 300°C, the dependence \(E_{\text{br}} = f(\tau)\) shows a minimum in electrical strength, which shifts toward shorter times (from 60 h to 15 h) as the aging temperature increases from 250 to 300°C.
Fig. 2. Dependence \(E_{\text{br}} = f(\tau)\) for polyimide film PM aged at a temperature of 250°C
It may also be noted that, as the aging temperature increases, the electrical strength decreases. Thus, the curve \(E_{\text{br}} = f(\tau)\) at 250°C lies lower than at a temperature of 200°C.
In order to determine the reason for the formation of the minimum in the dependence \(E_{\text{br}} = f(\tau)\), we studied the electrical strength of a polyimide film during thermal aging at a temperature of 250°C. The results of breakdown tests of the polyimide film, obtained by an analogous method, are presented in Fig. 2.
As statistical verification showed, the electrical strength of the polyimide film is practically independent of aging time over 250 h. This makes it possible to assume that the formation of the minimum in the curve of the dependence \(E_{\text{br}} = f(\tau)\) for Isoflex is associated with destruction of the polyepoxy binder during thermal aging.
Fig. 3. Dependence of the apparent intensity of partial discharges \(Q\) on aging time at T = 300°C for Isoflex
Indeed, as visual observations showed, during thermal aging Isoflex swells, followed by delamination. It may be assumed that, during thermal aging at temperatures of 250 and 300°C, thermal destruction of the binder occurs, accompanied by the release of gaseous products that lead to delamination. As the aging temperature is increased, the process of binder destruction proceeds more intensively and, for example, at an aging temperature of 300°C the process of delamination of Isoflex is observed—
visually after 5–10 hours of aging. When observed under a microscope, a large number of gas inclusions can be seen in the volume of the isoflex; their size increases during the process of thermal aging.
This phenomenon promotes redistribution of the electric field in the volume of the material and an increase in its nonuniformity as a result of partial discharges, which leads to a decrease in the electrical strength of isoflex.
To verify the assumption stated above, we studied the characteristics of partial discharges (p.d.) in isoflex during thermal aging at temperatures of 250 and 300°C. The results of measurements of the apparent intensity \(Q\) and the repetition rate of p.d., \(n\), obtained at an aging temperature of 300°C and a voltage of 6 kV, are presented in Figs. 3 and 4. Similar data were obtained at a temperature of 250°C.
As can be seen from Figs. 3 and 4, with increasing aging time the repetition rate and apparent intensity of p.d. initially increase, reaching maximum values at an aging time of 10–15 hours, and then decrease. From a comparison of Fig. 1 and Figs. 3 and 4 it is evident that the minimum electrical strength of isoflex at a temperature of 300°C corresponds, in aging time, to the maximum values of \(Q\) and \(n\).
Fig. 4. Dependence of the repetition rate of partial discharges \(n\) on aging time \(\tau\) at \(T = 300^\circ\text{C}\) for isoflex at different sensitivities of the PS-10000 pulse counter: \(I\)-1—\(n(1)\); \(II\)-2—\(n \cdot (10^{-1})\); \(III\)-3—\(n \cdot (10^{-2})\).
This makes it possible to conclude that the formation of gas cavities during thermal aging and the increase in their size lead to an increase in the apparent intensity of ionization, the repetition rate of p.d., and the nonuniformity of the electric field due to partial discharges. The latter causes a reduction in the electrical strength of isoflex.
With further aging, burnout of the binder occurs and the pores become shunted by destruction products. Indeed, measurements of the surface resistance of glass-varnish fabric impregnated with polyether-epoxy varnish showed that, during thermal aging, its value decreases by 3–4 orders of magnitude (from \(3.3 \cdot 10^9\) to \(7.8 \cdot 10^5\ \Omega\)). This leads to equalization of the electric field in the volume of the material, a decrease in the repetition rate and apparent intensity of p.d., and also to a certain increase in the electrical strength of isoflex (see Fig. 1).
Thus, the investigations carried out make it possible to draw the following conclusions:
-
The electrical strength of Isoflex at an aging temperature of 200°C is practically independent of the aging time over the course of 1000 hours. At aging temperatures of 250 and 300°C, a minimum is observed in the dependence of electrical strength on aging time.
-
The formation of a minimum in the dependence \(E_{\mathrm{br}} = f(\tau)\) is associated with destruction of the polyepoxy binder during thermal aging, leading to the formation of gas cavities and to an increase in the nonuniformity of the electric field due to partial discharges.
REFERENCES
-
Modern Composite Materials. Edited by A. Braushman and R. Kroon. Moscow, “Mir,” 1970.
-
V. Ya. Rizhkinashvili. Device for determining the electrical strength of solid dielectrics. “Plastmassy,” 1966, No. 10.
-
N. V. Smirnov, I. N. Dunin-Barkovsky. A Course in Probability Theory and Mathematical Statistics (for Engineering Applications). Moscow, “Nauka,” 1965.