HEATING AND COOLING OF END COMMUTATORS OF DIRECT-CURRENT ELECTRIC MACHINES

A. F. GRITSENKO

(Recommended by the seminar of the Departments of Electrical Machines and General Electrical Engineering)

Despite the considerable operational and economic advantages of end commutators \[1—5\], they have not yet found wide application, especially in large direct-current machines. Because of the limited height of the profile of the commutator copper in an end commutator, it is not always possible to obtain the same area of the contact surface and heat-dissipating surface as in a cylindrical commutator. It is necessary to increase the current density under the brush. This causes designers to have concerns about the reliability of the machine with respect to commutation and thermal loads. In \[6—9\] it is shown that an increase in current density is possible and even necessary in machines with cylindrical commutators for optimal operation of the sliding contact. As for end commutators, such a possibility of increasing current density from the standpoint of commutation reliability exists to an even greater extent \[4\]. However, the problems of heat transfer of an end commutator associated with an increase in current density have so far been little studied. In connection with this, the present work sets the task of investigating the cooling and heating of end commutators.

Cooling Conditions of End Commutators

Let us consider the two most widespread methods of commutator cooling—with an independent ventilation system and with self-ventilation of the electric machine. With an independent ventilation system, in most cases the cooling air enters the commutator chamber from the side, i.e., in the radial direction. To evaluate the degree of blowing of the commutator, the maximum air velocities near the contact surface were measured at the central point of each quadrant of the working surface of the commutator bounded by the brushes. The measurements were performed using a Pitot tube and a micromanometer on an experimental EDT-166 traction motor with an end commutator, with the armature stationary, and on the same traction motor with a cylindrical commutator. The measurement results are presented in Table 1.

The air entered the commutator chamber from the side of the 1st and 4th quadrants. The air flow rate in both cases was 40 m³/min.

Higher air velocities along the working surface of the end commutator are obtained owing to the fact that, during motion in the col-

Table 1

In the commutator chamber the cooling air encounters on its path only the brushes and the shaft. In an electric motor with a cylindrical commutator, the cooling air encounters on its path, in addition to the brushes, the brush holders and the cylindrical body of the commutator; as a result, the air velocity is reduced considerably, and from the side opposite the air inlet the cylindrical commutator is blown only weakly.

It should be noted that, in the experimental traction-motor specimen, the design of the commutator chamber had not yet been refined from the standpoint of obtaining high velocities of air blowing over the commutator; however, such a possibility exists by reducing the length of the commutator chamber in the axial direction.

With self-ventilation, the face commutator is also in more favorable conditions than the cylindrical one. Before entering the ventilation ducts, the supposedly cooling air passes along the working surface of the commutator. In a machine with a cylindrical commutator, the motion of the cooling air along the commutator plates in the axial direction is impeded, since the yoke, brush holders, and commutator cocks hinder this motion. The cylindrical commutator is cooled mainly by entrainment of air by the cylindrical surface during its rotation and by blowing air through the ventilation openings in the bushing.

Heat Transfer of the Face Commutator

Heat removal from the commutator takes place through the surface of the copper and of the commutator bushing, through the surface of the brushes and brush holders, and by heat conduction from the commutator to the shaft. In addition, a heat inflow is observed from the armature winding to the commutator through its connections with the commutator. It is very difficult to determine experimentally what quantity of heat, and along which path, is removed from or, conversely, supplied to the commutator. Therefore, in order to estimate the intensity of heat transfer of the face commutator, a comparison was made of the conditional heat-transfer coefficients of the face commutator and of the widely used and more thoroughly studied cylindrical commutator. In the present case, the conditional heat-transfer coefficient is understood as the ratio of the sum of the losses in the commutator due to brush friction and in the brush contact to the product of the mean temperature rise of the working surface of the commutator over the cooling air and the area of the copper surface in contact with the cooling air.

The experiment to determine the heat-transfer coefficients of the commutator under self-ventilation of the machine was carried out on the auxiliary generators of the two-machine unit VG/VGT of the TE-3 locomotive. On one generator a cylindrical commutator with micanite cuffs was installed; on another, exactly identical generator a face commutator on plastic, described in [5], was installed. The excitation on the generators was switched off; the commutator plates were short-circuited with one another; current was supplied to the brushes from a low-voltage direct-current generator. Identical ...

brushes and brush holders with identical pressing forces. During the experiment, identical rotational speeds and currents were established for analogous generator operating conditions.

The dependences of the conditional heat-transfer coefficients of commutators on the mean linear velocity of the working surface of the commutator are presented in Fig. 1: curve 1 is for the end commutator and curve 2 for the cylindrical one. From a comparison of the curves it can be seen that, at identical commutator temperatures, the heat loads on the end commutator may be increased approximately 1.5 times relative to the cylindrical commutator.

Higher values of the heat-transfer coefficients of the end commutator are explained, in addition to the reasons described earlier, also by the fact that in an end commutator the length of the cylindrical surface at the maximum diameter of the copper, owing to design features, is three times greater than the length at the same diameter of the cylindrical surface of the segments of a cylindrical commutator. With the shorter length of the ventilation ducts of the end commutator, a greater air flow through the armature is produced and, consequently, the commutator, relative to a generator with a cylindrical commutator having the same fan. The thin layer of plastic on the end commutator has high thermal conductivity from the copper to the commutator bushing in comparison with the air layers in the space where the micanite cylinder is located in a cylindrical commutator.

Fig. 1. Dependence of the heat-transfer coefficients on commutator speed under self-ventilation of the machine.
1 — end commutator.
2 — cylindrical commutator.

A large enhancement of heat transfer occurs through the use, in the end commutator, of a special fan for cooling the commutator, mounted on the bushing in the space between the shaft and the inner boundary of the copper plates.

Fig. 2. Dependence of the heat-transfer coefficients on the speed of the end commutator.
1 — with fan
2 — without fan.

In Fig. 2, curve 1 represents the dependence of the conditional heat-transfer coefficients on the mean linear velocity of the working surface—

the commutator velocity \(V_k\) and the average linear velocity of the fan blades \(V_v\) when a special fan is used. Curve 2 in Fig. 2 represents the heat-transfer coefficients of the end commutator without the use of any means of intensified cooling. The experiments for obtaining the curves in Fig. 2 were carried out on a separately mounted short-circuited end commutator of diameter 270 mm. On the side adjacent to the armature, the commutator was covered with heat-insulating material. In large direct-current machines, such, for example, as the MPT 99/47 traction generator with an armature diameter of 990 mm, a special fan for the commutator can be built in at low cost; with a linear velocity of 30–35 m/s, this will ensure intensive cooling of the commutator. With an independent ventilation system, the end commutator is likewise cooled intensively. Thus, for example, during thermal tests of the EDT-166 traction motor with an end commutator, the commutator temperature proved to be even somewhat lower than that of a traction motor with a cylindrical commutator under identical test conditions. For the reasons indicated above, the current density under the brushes in this traction motor with an end commutator was \(16.6\ \mathrm{A/cm^2}\), and the area of the working surface of the commutator was 1.5 times smaller than that of the cylindrical one.

It should also be noted that the overheating of the armature winding was somewhat lower in the motor with the end commutator, evidently because the reduction in the commutator length affected the aerodynamic resistance and the air flow through the armature.

Temperature distribution over the commutator surface

The nonuniform distribution of contact forces along the brush length [10], the nonuniform distribution of current under the brush, and also the difference in the cooling conditions of the surface of the end commutator were the reason for carrying out work to measure the temperature of the working surface of the commutator.

The experiments were carried out on a separately mounted end short-circuited commutator. The temperature was measured by copper–constantan thermocouples installed in the radial direction on the plane of the commutator, and also on the cylindrical part. The thermocouples were connected through slip rings to a potentiometer. Three thermal regimes were carried out with different methods of cooling the commutator: with independent air supply, with a fan specially built in for the commutator, and without any special means of intensified cooling. The results of the experiments showed that, despite the difference in the operating conditions of individual points of the working contact surface, the difference in temperatures in the radial direction, over the brush length of 44 mm, is no more than 1–3 degrees.

Such relatively good temperature equalization occurs because of the good thermal conductivity of copper. Thus there should be no fear of excessive thermal deformations of the commutator from nonuniform heating.

REFERENCES

1. A. E. Alekseev. Design of Electrical Machines. Gosenergoizdat, 1958.

2. M. F. Karasev. Design of Commutator Direct-Current Machines. Gosenergoizdat, 1961.

3. I. I. Tuktaev. Dynamics of sliding contact. Izvestiya vuzov. Elektromekhanika, 1959, No. 11.

4. V. I. Nallin, I. I. Tuktaev, N. D. Zharkov. Application

end commutators and the study of the operation of a flat sliding contact. Proceedings of VNIEM, vol. XIX, Energiya Publishing House, 1965.

1. A. F. Grishchenko. End commutators for large direct-current machines. Elektromekhanika, 1966, No. 9.

2. T. G. Abartsumov, E. M. Kovarskii, G. I. Gershnovich. On the possibility of increasing the permissible current density under brushes. Vestnik elektropromyshlennosti, 1958, No. 10, 1.

3. V. P. Tolkunov. On choosing the value of the current density under brushes in direct-current machines of small and medium power. Elektrichestvo, 1962, No. 8.

4. P. S. Livshits. Optimum operating conditions of an electrical sliding contact. Vestnik elektropromyshlennosti, 1962, No. 8.

5. A. Ya. Gluskin, A. M. Bordachenkov, V. P. Stepanov. Operation of the brush contact of electric machines at increased current density. Elektromekhanika, 1963, No. 9.

6. A. F. Grishchenko. Brush wear on an end commutator. Theory and Practice of Commutator Machines. Transport Publishing House, 1965.