Study of Heating in AO-Type Asynchronous Motors

V. A. Zhadan, Yu. V. Kopylov, D. I. Sannikov

(Presented by the scientific seminar of the departments of electrical machines and general electrical engineering)

The study of heating in enclosed, fan-cooled asynchronous motors was carried out on motors of the currently produced AO2 series and on experimental samples of motors of the D and AO4 series, having a center height of 132 mm, under rated operating conditions at steady-state heating.

The temperature field of the motors under test was determined by means of copper-constantan thermocouples embedded in various parts of the motors. In the stator winding 40–60 thermocouples were embedded; in the rotor, 10–20 thermocouples; in the AO2-42-4 motor, 12 thermocouples were embedded, making it possible to measure the temperature difference in the slot insulation; the overheating of the surface of the stator back was determined in its middle part by 12 thermocouples uniformly distributed around the circumference. The overheating of the frame and bearing shields was measured by 64 thermocouples. The readings of the rotor thermocouples were measured, with the motor rotating, through a slip-ring device. The locations of thermocouple embedding and the temperature fields of the corresponding motors are given in Figs. 1–3. Here the overheating values of the rotor, frame, and bearing shields are averaged over the circumference; the overheating values of the stator winding are averaged over the circumference of the stator and over the height of the slot.

An explanation of the character of the curves and the calculation of parameters characterizing the temperature scatter among individual thermocouples embedded in the stator winding and in the frame are given in [2].

The hottest part of the motor is the rotor. The temperature along the length and circumference of the rotor varies little. On average, the rotor of 4-pole machines is hotter than the stator by 25–30°C. The temperature drop in the slot insulation varies along the length from 5 to 8°C; however, the thermocouples measuring the temperature of the outer surface of the insulation are installed in such a way that their heads may be pressed into the insulation. Therefore the temperature drop in the slot insulation is in reality 1.15–1.25 times greater.

In [2] it was noted that the heads of the end-winding portions constitute a monolithic structure that has impeded cooling conditions. The installation of spacing pads between sections allows air to pass freely through the channels thereby formed and improves heat transfer from the end-winding portions. Thus, in the AO2-42-4 motor their average overheating decreases as a result by 2.8°C, or 4.4%; the overheating of the slot portion of the stator winding decreases only by 0.8°C, or 1.4%.

This shows that improvement of heat transfer from the end-winding portions only slightly reduces the temperature of the winding as a whole.

If, on the contrary, the thermal resistance from the end-winding portions is increased—

![Graph and cross-section diagram. Curve labels: \(\theta\), \(\theta_p\), \(\theta_{M1}\), \(\theta_{u3}\), \(\theta_{ct}\). Legend: \(\triangle\ \circ\ \nabla\ \square\) AO2-42-4; \(\bullet\ \blacksquare\) AO2-42-2.]

Fig. 1. Temperature field of motors AO2-42-4 and AO2-42-2

...to the internal air—for example, by coating them with asbestos—their average temperature increases by \(4^\circ\)C, and that of the slot portion by \(3.6^\circ\)C. At the same time, the temperature drop from the end portions to the slot portion increases from \(5.2\) to \(5.6^\circ\)C. Consequently, the heat flow from the end portions into the slot portion increases only slightly. Thus, the deterioration of heat removal from the end portions to the internal air does not play a significant role. The rise in the temperature of the stator winding in this experiment is explained by the following reason: the additional insulation on the end portions impeded air circulation and worsened heat transfer from the rotor; its temperature rose by \(8^\circ\)C, and the fraction of heat transmitted from the rotor to the stator through the air gap increased, which caused additional heating of the slot portion of the winding.

Air circulation inside the motor is provided by an impeller cast integrally with the rotor short-circuiting ring. It is known that the intensity of air mixing depends on the size and number of blades. Since the tests were carried out on a production AO2-42-4 motor, in this case it was possible to vary the width of the impeller...

blades only in the direction of reduction from the initial value of 46 mm; the height and number of vanes were not changed. When the width of the impeller was reduced by a factor of 2, the average overheating of the stator winding increased by 3.8%, and the overheating of the rotor increased by 5.9%; with the impeller completely stopped, the average overheating of the stator winding increased by 11.6%, and the overheating of the rotor by 15%. The increase in rotor overheating in this case occurs as a result of the reduction in the cooling surface of the impeller, and that of the stator is due to the increase in heat flow from the rotor through the gap and the deterioration of the blowing of the end parts. Although reducing the width of the impeller reduces the internal ventilation losses, their effect on the thermal state of the motor is insignificant in comparison with other types of losses. It was not possible to find the optimum width of the vanes, as was shown in [1]; apparently it is greater than in the existing motor design.

The design of the internal fan formed by the impeller is such that the share of air flows circulating along leakage paths and not participating in the cooling of the end parts of the stator winding is significant. In addition, between the stator and the frame there are places where practically no air mixing occurs. Therefore, in order to regularize the movement of air inside the machine, it was proposed to install a disk in the leakage path, i.e., to make the fan a two-disk one and thereby direct the air into the lattice formed by the end parts at the slot exit. As a result, the overheating of the end parts of the stator winding decreased by 1.5°C, and the average overheating of the stator winding by 0.6°C, or by 1%.

A very important thermal parameter for enclosed motors is the conductance between the back of the stator core and the frame, which depends to a large extent on the condition of the contacting surfaces, contact pressure, etc. Studies on various machines show that the temperature drop in the gap between the core and the frame can range from 3 to 7°C.

In the AO2 series the stator core is not machined on the outside diameter, and the average drop between the core and the frame is 5.7°C. To determine how the machining class affects the temperature drop, six motors were manufactured and tested in which the stator core was machined to class ▽ 6. The temperature drop was 5.4°C, i.e., remained practically unchanged. Apparently, stricter requirements for the machining of the seating surfaces of the stator core and the frame do not ensure a substantial reduction in the temperature drop between them.

The character of the temperature field (Fig. 2) of the experimental motor sample D—132 M-4 is the same as that of the AO2-series motor. In general, the temperature of the stator winding and frame is higher than in motors of the AO2 series.

The temperature field of the first experimental samples of the AO4 series motors is shown in Fig. 3.

The hottest points of the stator winding are located in the end parts on the side opposite the fan. The increase in the temperature of the hot point over the average winding temperature is composed of two components:

\[ \Delta \theta_{\mathrm{M}} = \partial \theta_{\mathrm{M1}} + \partial \theta_{\mathrm{M2}}, \tag{1} \]

where \(\partial \theta_{\mathrm{M1}}\) is the excess of the half-sum of the temperatures of the ends of the end parts over the average winding temperature, caused by the influx of heat from the end parts to the slot part.

\(\partial \theta_{\mathrm{M2}}\) is the skewing of the winding temperature field due to asymmetry in cooling of the housing, i.e., the half-difference of the temperatures of the end parts.

Fig. 2. Temperature field of motor D—132M-4

The first component can be calculated, under the condition of symmetrical cooling of the stator, on the basis of the assumption of a parabolic character of the temperature distribution along the length of the end-winding part. Omitting the mathematical derivations, in final form we have

\[ \delta \theta_{M1} = P_{M1} k_{\ell} \frac{1}{24 S_M \lambda_{\mathrm{Cu}}} \cdot \frac{l_{\ell}^{2}(l_{\ell}+3l)}{(l_{\ell}+l)^2}. \tag{2} \]

Here \(P_{M1}\) are the losses in the copper of the stator winding,

\[ k_{\ell} \]

is 0.75 for two-pole machines and 0.95 for the others,

\[ S_M \]

is the total copper cross section;

\[ \lambda_{\mathrm{Cu}} \]

is the thermal conductivity of copper;

\(l_{\ell}\) and \(l\) are the lengths of the end-winding and active parts of the stator winding.

The second component is determined through the skewing of the temperature field of the frame, taking into account the thermal resistance between the winding and the frame and the equalizing action of the axial thermal conductivity of the winding.

The skew coefficient of heat transfer of the housing [3] is recommended to be calculated by a formula that takes into account the length of the frame,

\[ k = 0.92 + 0.7\left(\frac{L_{\mathrm{st}}}{D_{\mathrm{st}}} - 1\right), \tag{3} \]

Fig. 3. Temperature field of motors AO4-132M-4 and AO4-132M-6

where \(L_{\mathrm{st}}\) and \(D_{\mathrm{st}}\) are the length and outside diameter of the frame.

The skew of the temperature field of the frame, equal to half the difference of the temperatures at its edges, is calculated by the simplified formula [3]

\[ \Delta \tau = \theta_{0}\,\frac{0.4\,k}{1+\dfrac{10}{R_{\mathrm{st}}G_{\alpha}}}, \tag{4} \]

where \(R_{\mathrm{st}}\) is the axial thermal resistance of the frame,
\(G_{\alpha}\) is the thermal conductance from the shell to the air.

The component of the hot-spot temperature rise is

\[ \delta \theta_{\mathrm{m}2} = \Delta \tau\, \frac{ \dfrac{1}{2}\left(1+\dfrac{l}{L_{\mathrm{st}}}\right) }{ 1+\dfrac{12R_{\mathrm{ms}}}{R_{\mathrm{m}}} }. \tag{5} \]

Here

\[ R_{\mathrm{m}}=\frac{l+\dfrac{2}{3}l_{\mathrm{p}}}{S_{\mathrm{m}}\lambda_{\mathrm{Cu}}} \]

— the calculated axial thermal resistance of the stator winding.
\(R_{\mathrm{mc}}\) is the thermal resistance between the winding and the frame, consisting of the resistances of the insulation, the tooth over \(1/3\) of its height, the stator back, and the gap between the stator core and the frame.

The excess of the temperature of the hottest point over the average winding temperature for the AO2-42-4 motor, according to experiment (Fig. 1), is \(4.5^\circ\mathrm{C}\), and according to calculation, \(4.6^\circ\mathrm{C}\); for the AO2-42-2, according to experiment, \(5.7^\circ\mathrm{C}\), and according to calculation, \(6.1^\circ\mathrm{C}\).

Thus, the excess of the temperature of the hottest point over the average temperature of the stator winding, which characterizes the nonuniformity of heating of the stator winding along its length, has a significant value (up to 10%).

The proposed method for calculating the excess temperature of the hottest point over the average winding temperature has sufficient accuracy.

REFERENCES

1. F. L. Zautner, I. I. Feigelman, M. M. Andzheiko, V. I. Borisovich. Optimal length of the ventilation blades of the squirrel-cage rotor of explosion-proof asynchronous motors. Elektrotekhnika, 1965, No. 7.

2. V. A. Zhadan, D. I. Sannikov, R. Ya. Klyai. Study of temperature fields of enclosed ventilated electric motors. Izvestiya TPI, vol. 172, 1967.

3. V. A. Zhadan, D. I. Sannikov. Asymmetry of heating of the frame of externally ventilated electrical machines. Izvestiya TPI, vol. 172, 1967.