METHOD FOR MODELING THE COMMUTATOR PROFILE TO STUDY THE EFFECT OF THE CONDITION OF THE COMMUTATOR SURFACE ON THE ELECTRICAL AND MECHANICAL CHARACTERISTICS OF BRUSH CONTACT

A. I. Skorospelshkin, L. Ya. Zinner, E. K. Damm

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

It is known that the mechanical condition of the brush–commutator assembly has a great, and in some cases decisive, influence on the commutation of commutator machines. Of the totality of mechanical factors affecting commutation, the principal one is the commutator profile; therefore, the importance of studying the effect of the commutator profile on the mechanical behavior and on the electrical characteristics of the sliding electrical contact is obvious.

The study of the influence of the commutator profile on current collection, commutation, and machine reliability encounters certain difficulties because, first, it is difficult to select a machine whose commutator profile would be sufficiently typical, and, second, when selecting various machines, it is impossible to vary the commutator-profile parameters according to a preassigned program.

Therefore it becomes necessary to model artificially the commutator profile of both real machines and artificial apparatus for studying commutation.

Since the actual steps between commutator bars after mechanical machining of commutators are small and amount to 1–5 microns, it is clear that the required commutator profile cannot be reproduced with sufficient accuracy by generally available means of mechanical machining. In this case the most acceptable method is galvanic coating of the commutator surface.

Figures 1 and 2 (a, b) show the planned profiles of two commutators of series P machines of the third size and the profilogram of one of these commutators after treatment in a galvanic bath. The oscillogram of the profile was obtained by means of a profilometer with a capacitive transducer [1].

The method for reproducing the required profile, as applied, for example, to the commutator of Fig. 1, is as follows. The first commutator bar and all metal surfaces (shaft, bearings, commut—

Fig. 1. Planned commutator profile

Fig. 2. a — planned commutator profile; b — realized profile

...the end bushing, etc.) are coated with paraffin. The remaining commutator bars are carefully degreased. All commutator bars must be short-circuited (if they are not short-circuited by the armature winding). The system is lowered into a bath with a sulfuric-acid solution. The electrodes from which the copper is transferred should, as far as possible, be uniformly arranged around the commutator. Thus, the current referred to a unit of the surface being plated should be chosen small (up to \(0.01\ \text{A}/\text{cm}^2\)) in order to avoid nonuniformity of the coating. From the known current density, the time is calculated during which the coating thickness will be 2 microns. After this time has elapsed, the current is switched off and, for example, bar No. 6 is additionally closed, etc.

Analysis of the oscillograms shows that the accuracy with which the required profile is realized is quite sufficient.

The commutators whose profilograms are shown in Figs. 1 and 2 were intended for studying radial brush vibrations and for determining the effect of steps between commutator bars on the natural dynamic current-voltage characteristics of the brush–commutator-segment contact. Commutator models made by the method described can also be used for other investigations.

As is known, the static and dynamic current-voltage characteristics of a brush contact depend to a large extent on the commutator profile. When taking these characteristics for the purpose of their further approximation and use in commutation calculations, it is necessary to have a commutator whose profile has been designed and reproduced from the results of a statistical study of profilograms of commutators of machines of the given type.

The principal parameters of a commutator profilogram are:

\(\sigma\) — the root-mean-square step between segments,

\(e\) — eccentricity,

\(E\) — ellipticity of the surface.

A statistical study of a sample from a batch of commutators of the giv-

of this type makes it possible to find the most characteristic values of $\sigma$, $\varepsilon$, and $E$ and to create a statistical model of the commutator by the method described.

Measurement of brush vibrations and of the contact voltage drop between the brushes and individual segments on the statistical model of the commutator will also make it possible to draw a conclusion as to the extent to which the commutator profile produced by a given technology for machining its surface ensures stable operation of the brush contact in mechanical and electrical terms.

In addition, modeling the commutator profile can be used to investigate the nonidentity of commutation cycles of DC machines caused by mechanical factors, to develop a system of tolerances for the parameters of the commutator profile, and for experimental verification of various theories of brush vibrations.

REFERENCES

1. L. Ya. Zinner, A. I. Skorospeshkin. A bridge instrument for studying the dynamics of commutators of electrical machines. Izvestiya TPI, vol. 160, 1966.