A New Methodology for Preparing and Conducting Tests of Stators and Armatures of Electrical Machines in the Study of Winding Insulation Durability

A. Ya. Tsirulik, V. N. Popov

(Presented by the scientific seminar of the Department of Electrical Machines)

In order to obtain reliable information on the durability of windings of electrical machines, it is necessary to test a sufficiently large number of machines if the traditional principle is applied—conducting the test until winding failure. The tests prove to be lengthy and expensive. In [1] a methodology was proposed, and is currently used for testing asynchronous motors with power up to 100 kW, which makes it possible, without bringing the machine to failure under accelerated tests, to evaluate changes in the properties of the insulation in comparison with the initial state by repeatedly testing, on each machine, the ground and turn-to-turn insulation with high voltage to breakdown. Repeated determination of the magnitude of the breakdown voltage of the insulation on one machine makes it possible to obtain more information about the state of the insulation, thereby making it possible to reduce the number of machines tested and to lower the cost of the tests. The proposed methodology is not without shortcomings, which consist in the fact that, in order to test the turn-to-turn insulation with voltage after conducting durability tests of the machine, it is necessary to carry out a number of labor-intensive operations:

1. Before placing the sections of a random-wound winding into the slots, it is necessary to tie the turns of the sections together in pairs with thread.

2. After conducting durability tests of the machines, it is necessary to cut any end-winding part of the winding on one side of the core, separate the conductors, locate and fix 100–150 pairs of conductors tied together in pairs, cut off the excess conductors, impregnate with varnish, and dry the end-winding part of the machine prepared in this way. After this, the insulation can be tested by applying voltage between pairs of conductors.

Thus, the disadvantage of the methodology consists in its laboriousness and in the fact that one end-winding part of the winding is destroyed and does not take part in the testing of the winding insulation.

In the present article a new methodology is proposed, free of the indicated shortcomings. Its essence is as follows. Before the winding sections are placed into the slots of the stator or rotor, test conductors (Fig. 1) or turns (Fig. 2) are placed in the sections and fastened to the working turns of the section with threads at the places indicated in the figures by crosses. The test conductors or turns are made from the same winding wire as the winding sections. The ends of the test turns must protrude from any part of the section on the side opposite to the ends of the section. One end of the test turn should

Fig. 1. Fastening the test conductor in the stator-winding section: 1 — test conductor, 2 — winding section

Fig. 2. Fastening the test turn in the armature-winding section: 1 — test turn, 2 — winding section.

...be made short. In each section it is necessary to place such a number of test turns or conductors that their total number in the entire stator (armature) is no less than 70–100, i.e., no less than the minimum required number of points for determining the parameters of the distribution law of the breakdown voltages of the turn-to-turn insulation. It is desirable that in each section there be no more than 4–5 test conductors per side, or turns. In order for the slot fill factor to remain unchanged, the winding and the test turns must be made from winding wire whose diameter is determined taking into account the increased number of conductors in the slot. If there are more than 30 turns in a section, the wire diameter need not be changed, since 5–6 test conductors inserted into the slot change the fill factor only slightly.

Sections with test turns or conductors placed in them are laid in the slots of the stator or rotor in the usual manner. In armature windings of direct-current machines, test turns should be used, while in stator windings test conductors may be used, since this makes it possible to reduce by half the required number of conductors.

After the stator winding has been laid, the ends of the test conductors protruding from the end portions must be bent onto the end portion of the winding, arranged uniformly around the circumference on the outer surface of the winding, and bandaged. The end portion must first be insulated. On the armatures of direct-current machines, the protruding ends of the test turns on the side opposite the commutator are bent onto the end portion of the winding, previously insulated, after which insulation is applied on top and the armature is bandaged. Impregnation and drying of stators and armatures are carried out in the usual manner...

then the machines are assembled and subjected to accelerated life tests. After the tests are completed, the machines are disassembled, and the stators (armatures) are prepared for testing the turn-to-turn insulation of the windings by voltage; for this purpose the end portions of the windings, to which the ends of the test conductors are fastened, are unbanded and the conductors are lifted. No damage to the winding insulation occurs in this process, since during manufacture of the winding the conductors were separated from the winding by additional insulation. To test the turn-to-turn insulation, voltage must be applied in succession between the winding and each test conductor; in this case two layers of turn-to-turn insulation are subjected to the test, as in any other test method. One layer is the insulation of the test conductor; the other layer is the insulation of the active conductors of the winding with which the test conductor is in contact. The test is carried out over a conductor length equal to the length of that part of the test conductor or turn that lies within the boundaries of the winding.

The advantage of the proposed methodology is that preparing stators or armatures for testing turn-to-turn insulation is very simple, does not require destruction of the end portion of the winding, and therefore makes it possible to obtain more reliable data on the condition of the insulation after accelerated life tests of machines. The present methodology is used by the authors in conducting investigations of the life of series “P” direct-current machines at the Elektromashina plant (Prokopyevsk) and at Tomsk Polytechnic Institute.

REFERENCES

1. A. P. Matyalis, E. K. Strelbitskii. A method for investigating aging of the insulation of low-voltage electrical machines. Izv. TPI, vol. 242.