Study of the Combined Effect of Temperature and Vibration on Defects in Turn Insulation of Asynchronous Motors

Yu. P. Pokhodkov, P. P. Besperstov, V. S. Klyuev,
V. V. Pykhtin

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

In the course of operation of electrical machines, the insulation of their windings is subjected to the influence of a whole series of factors: heat, vibrations, electrodynamic forces, moisture, commutation overvoltages, etc. Depending on the level and simultaneity of these factors, the degree of their influence on the insulation varies.

In the present work, a study was carried out of the simultaneous influence of temperature and vibrations on the turn insulation of the windings of asynchronous motors. As the object of the investigations, the winding of an asynchronous motor was taken, made with PETV-\(F\)-35 wire of diameter \(1.08\) mm.

The investigation was conducted on mock-ups representing statorettes. Pairwise-connected conductors were placed in the slots of stator packs. On one side the end portion of the winding was cut, and the ends of the conductors were stripped.

The laying of the winding in the stator slots, impregnation, and drying of the windings were carried out in full accordance with the manufacturing technology for the winding of an AO2-32-4 motor. The slot fill factor was 0.82. ML-92 lacquer was used as the impregnating compound. To ensure the simultaneous effect of temperature and vibration on the winding, the winding mock-ups were placed in a thermostat mounted on an ST-1000 vibration stand.

Table 1 gives the levels of the acting factors and the testing time for the insulation specimens.

Table 1

Note: \(g\) is the acceleration of free fall, \(g = 9.8\ \text{m/sec}^2\); vibration frequency 50 Hz.

As the criterion for evaluating the quality of the insulation, the value of defectiveness was adopted—the probability of the occurrence of pairs of conductors with insulation,

the breakdown voltage of which is equal to or less than the maximum breakdown voltage of air in a thickness equal to twice the thickness of the wire insulation, taking into account surface flashover along the insulation over a distance of no more than 3 mm. According to experimental data, this voltage is 4.8 kV; consequently, the defectiveness \(q\) was determined from the expression

\[ q=\frac{n_{4,8}}{\sum n_i}, \tag{1} \]

where \(n_{4,8}\) is the number of conductor pairs broken down at \(U\leq 4.8\) kV, and \(\sum n_i\) is the total number of tested specimens. The defectiveness of the wire insulation after each experiment was determined from the results of breakdown of the insulation of 90–100 pairs of conductors of the model at alternating voltage, the rate of rise of which was 1 kV/sec.

Analysis of the distribution of the breakdown voltage of the insulation of conductor pairs that had not been subjected to exposures (Fig. 1, curve 1) indicates the absence of through, coincident defects.

Fig. 1. Dependence of \(\sum P_i\) on \(U_i\) at \(t=220^\circ\)C: 1 — \(a=0g\) and \(t=20^\circ\)C; 2 — \(a=0g\); 3 — \(a=3g\); 4 — \(a=6g\); 5 — \(a=9g\)

The distribution of the breakdown voltage of the insulation of conductor pairs after exposure indicates an increase in the number of defects in the insulation (Fig. 1, curves 2, 3, 4, 5). The presence of two breaks in the distribution of \(U_{\mathrm{br}}\) can apparently be explained by the presence of three types of insulation compositions of two adjacent conductors.

The first type—the insulation of both conductors has through cracks lying at a distance of \(< 3\) mm from one another. The breakdown voltage of such pairs is \(\le 4.8\) kV.

The second type—the insulation of only one of the two conductors has through damage. The breakdown voltage of such a composition lies within the range \(4.8 \div 9.6\) kV.

The third type—the insulation of the conductors has no through defects. The breakdown voltage of a composition of this type is \(> 9.6\) kV.

Figure 2 shows the dependence of the defectiveness of the insulation, found from formula (1), on the magnitude of the vibration acceleration. The level of defectiveness

Fig. 2. Dependence of the defectiveness of turn insulation on the magnitude of vibration acceleration: 1 — \(200^\circ\)C; 2 — \(220^\circ\)C

of the insulation increases both with increasing temperature and with increasing acceleration during vibration. Analysis of the results shows that at elevated temperature the influence of vibration is more substantial. At a temperature of \(220^\circ\)C, an increase in the level of defectiveness is observed at a vibration acceleration of \(3g\). In the case where the temperature is \(200^\circ\)C, significant growth of defectiveness is observed at accelerations of \((4.5+5)g\). The latter may be explained by the increased probability of rupture of chemical bonds at elevated temperature and by more intensive defect formation.

In the work, the rate of growth of insulation defectiveness \((V)\) was also determined.

\[ V = \frac{q_i - q_{20}}{t}, \]

where

\(q_{20}\)—defectiveness of the insulation in the initial state at \(t = 20^\circ\) and \(a = 0g\),

\(q_i\)—level of insulation defectiveness after the simultaneous action of temperature and vibration.

Table 2

Table 2 gives the values of defect growth rates at various temperatures and vibration accelerations.

Conclusions

1. The defectiveness of the insulation increases by a factor of 2.5–3 as the vibration acceleration increases from 0 to 9 g.

2. It has been established that vibrations have little effect on insulation defectiveness at accelerations below 3 g.

3. The acceleration level under vibration has a substantial effect on the defect growth rate. Without vibration, the defect growth rate is: at 220°—0.00487 1/hour; at 200°—0.003 1/hour.

At a vibration acceleration of \(a = 9\) g, the defect growth rate is: at 220°—0.0119 1/hour; at 200°—0.0088 1/hour.

LITERATURE

1. A. K. Mitropolsky. Technique of Statistical Computations. “Nauka,” 1971.

2. L. M. Bernstein. Insulation of General-Purpose Electrical Machines. “Energiya,” 1965.