Effect of the Impregnating Composition on the Reliability of Turn Insulation in Induction Motors

Yu. P. Pokhodkov, V. I. Derevyanko

(Presented by the scientific seminar of the Department of EIKT)

The growth in the production of electric motors leads to a constant increase in the absolute number of motors that fail before the end of the warranty period. Therefore, the study of reliability and the identification of the reasons for its improvement is an important task of electrical-machine engineering. According to statistics, about 90% of motors fail because of failure of the turn insulation of windings, as a result of a decrease in the average breakdown voltage of the turn insulation of windings during operation due to the influence of many factors: temperature, moisture, commutation overvoltages, vibrations, etc.

The reliability of insulation is determined by the properties of electrical-insulating materials, which also include the impregnating composition. The impregnating composition slows the processes of thermal aging and moistening of the insulation, fills defects in the varnish coating of wires, etc., which should be accompanied by an increase in the reliability of the insulation. At present, varnishes 321-T, ML-92, and K-47K are used for impregnating electric-motor windings.

The main purpose of this work was to study the effect of the type of impregnating varnish, when using different diameters of winding wires and different slot-fill factors, on the reliability of the turn insulation of induction motors. In carrying out the experiments, the principal task was to make the conditions for testing the impregnating compositions correspond to the operating conditions of these compositions in electrical machines. The tests were conducted on a model simulating the arrangement of a pair of enameled wires in a motor slot (Fig. 1) and consisting of two parallel textolite plates, between which 25 pairs of parallel conductors, positioned opposite one another, were fastened at equal distances. Each pair of conductors had artificial insulation damage in the form of cuts down to the copper around the entire circumference; the damages were located opposite one another. The distance between the conductors was varied and fixed with the aid of Lavsan spacers 0.05 mm thick and having slots opposite the defects, which corresponded to a change in the slot-fill factor.

Before breakdown testing, the fixed conductors were impregnated with one of three varnishes: 321-T, ML-92, K-47K. Impregnation and drying were carried out in accordance with the technology adopted at the plant

“Sibelektromotor” for motors of the AO2 series. Breakdown tests were carried out on a high-voltage installation at industrial frequency. The specimens were made from PETV-f-35 wire, \(d_{\mathrm{iz}} = 1.64\) mm.

To determine the dependence of the electrical strength of impregnating compositions on the diameters of enamel wires, the following procedure was adopted. Artificial damage in the form of annular cuts down to the copper was first made on wires with insulation diameters of 0.83 mm, 1.16 mm, and 1.28 mm. Impregnation and baking of the specimens were carried out according to factory technology. Tests to determine the electrical strength of the varnishes were carried out on a high-voltage installation using a flat metal electrode, which corresponds to breakdown of two tightly adjacent conductors in the slot of a motor.

Fig. 1. Diagram of the device for determining the dependence of \(E_{\mathrm{pr}}\) on the slot filling factor

The test results are summarized in Table 1. It turned out that, as the diameter of the enamel wire increases, the average breakdown voltage per 1 mm of thickness of the impregnating layer decreases. This is explained by the fact that, with increasing wire diameter, the tested fraction of the annular cut increases, which leads to an increased probability of detecting defects.

Table 1

Dependence of \(E_{\mathrm{pr}}\) of the impregnating composition on the type of varnish and wire diameter, kV/mm

When considering the dependence of the average breakdown voltage of the impregnating layer on the distance between wires (Table 2), a tendency was observed for \(\overline{U}_{\mathrm{pr}}\) to decrease as the distances increased; this is most sharply expressed at distances from 0 mm to 0.1 mm for varnishes ML-92 and 321-T. This is explained by the fact that, at distances between conductors not exceeding \(0.2d_{\mathrm{iz}}\), varnish “bridges” are formed between the conductors, practically completely filled with varnish [2].

With an increase in the distances between conductors, the probability that “bridges” will appear decreases; the conductors are coated with varnish separately, which leads to a reduction in the average breakdown voltage of the gap. At distances between conductors from 0,1 mm to 0,2 mm, \(\overline{U}_{\mathrm{br}}\) changes insignificantly, since, apparently, varnish “bridges” are not formed; \(\overline{U}_{\mathrm{br}}\) will be characterized mainly by the thickness of the impregnating-varnish layer on the enameled wire, and an increase in the air gap between conductors does not lead to a substantial change in \(\overline{U}_{\mathrm{br}}\) of the gap. The different degree of dependence of \(U_{\mathrm{br}}\) on the distances between conductors is explained by the different viscosity of the impregnating varnishes, since as viscosity decreases the ability to form “bridges” decreases; therefore, for distances at which “bridges” are not formed, the breakdown voltage is characterized by the thickness of the varnish films covering the conductors.

Table 2

Dependence of \(\overline{U}_{\mathrm{br}}\) of the impregnating composition on the type of varnish and the distance between conductors, kV

Based on the test results, an attempt was made to find the degree of “non-impregnation,” which can be determined as the ratio of the number of regions having defects but not filled with varnish during impregnation to the total number of defective regions in the motor winding. In this case the impregnation substantially “heals” the defective regions; that is, the breakdown voltage of an impregnated region with a defect approaches in magnitude the breakdown voltage of regions without defects.

The degree of “non-impregnation” was determined for conductors \(d_{\mathrm{ins}} = 1,64\) mm at distances between them equal to 0,2 mm, and was expressed as the ratio of the number of conductors broken down by the maximum voltage of the air gap between conductors to the total number of conductors tested:

\[ p_{\mathrm{ni}}=\frac{n_{\mathrm{ins}}+3\sigma_{v}}{\sum_{i=1}^{n} n_i}; \qquad U_v+3\sigma_v \simeq 2,5 \text{ kV} \]

The results of the calculation are summarized in Table 3.

Table 3

Dependence of the degree of “non-impregnation” and the reliability of turn insulation (at \(S = 0,2\) mm) on the type of varnish and the distance between conductors \((S,\ \mathrm{mm})\)

* Theoretical reliability of the turn insulation of an asynchronous motor AO2-32-B as a function of the degree of “non-impregnation.”

The calculation of the reliability of turn insulation taking into account the degree of “non-impregnation,” carried out for the AO2-32-6 motor according to [3], showed that the motor impregnated with varnish 321-T has the highest reliability, followed by motors impregnated with varnishes K-47K and ML-92.

Conclusions

1. A method is proposed for determining the electrical strength of impregnating compositions as a function of the diameters of winding wires and the distances between them. The test results obtained by the proposed method make it possible to assess the degree of “non-impregnation” of defects in enameled-wire insulation and can be used to assess the reliability of the turn insulation of induction-motor windings.

2. The reliability of the turn insulation of a machine winding has been assessed taking into account the degree of “non-impregnation” of the gap between conductors in the machine slot.

3. An evaluation is given for three varnishes: 321-T, ML-92, K-47K. The windings impregnated with varnish 321-T have the highest reliability during the run-in period.

References

1. L. M. Bernshtein. Insulation of General-Purpose Electrical Machines. Energiya, 1965.

2. L. M. Bernshtein. Drying, Impregnation, and Compounding of Electrical-Machine Windings. Energiya, 1967.