Investigation of Damage to the Fibrous Insulation of Winding Wire During the Manufacture of Random-Wound Windings
Abstract Generated abstract
This paper examines whether manufacturing operations for random-wound windings damage the fibrous insulation of PSDKT winding wire and how such damage affects turn-to-turn breakdown voltage. The authors test batches of wire specimens with artificial scratches and cuts, with and without varnish impregnation, and compare these results with specimens taken from armatures before and after winding operations. The experiments indicate that mechanical damage far exceeding normal production damage has little effect on unimpregnated fibrous insulation, while impregnation mainly forms a surface film rather than filling insulation pores. The study concludes that PSDKT fibrous insulation is relatively insensitive to winding technology damage, and that as-delivered wire can be used for defectiveness data in reliability calculations, though the concept of defectiveness for porous fibrous insulation requires further clarification.
Full Text
STUDY OF DAMAGE TO THE FIBROUS INSULATION OF WINDING WIRE DURING THE MANUFACTURE OF RANDOM-WOUND WINDINGS
I. A. Volkomirskii, A. Ya. Tsirulin
(Presented by the scientific seminar of the Department of Electrical Machines)
The method for calculating the reliability of the insulation of random-wound windings of asynchronous motors [1] is based on the concept of an insulation “defect.” An insulation defect is understood as an elementary section of insulation whose breakdown voltage corresponds to the breakdown voltage over a length equal to the insulation thickness (cracks in the wire insulation, scratches, etc.). Defectiveness is the number of defects per unit length or area of insulation. A defect is present in the as-delivered condition, arises during the process of manufacturing the winding, and is formed as a result of aging during operation of the machine.
For wires with enamel insulation, the influence of winding-manufacturing technology on defect formation has been studied sufficiently well. It is of interest to determine and quantitatively evaluate the damageability of fibrous insulation of the PSDKT grade, widely used in electrical-machine building. An experiment was carried out in which five batches of conductors, 150 mm long, with 200 pieces in each batch, cut from PSDKT-grade wire taken from one coil, were tested. The conductors of the first batch had no insulation damage. The insulation of the conductors of the second batch was damaged artificially by making a longitudinal scratch with a knife along the entire length of the conductor. The insulation of the third batch of conductors was damaged by transverse cuts down to the copper. These damages imitate technological insulation damage during laying of the winding into the armature slots. To determine the degree of impregnation of the fibrous insulation by impregnating varnish, the next two batches of conductors were impregnated in K-47 varnish and dried. After impregnation, before drying, all conductors were carefully wiped with rags to remove the surface film of varnish, since the purpose of the experiment was to determine whether the varnish penetrates into the pores of the insulation and makes it monolithic. After drying, the fifth batch of conductors was damaged by making a longitudinal scratch.
Each batch of conductors was tested for turn-to-turn insulation with an alternating voltage of frequency 50 1/sec in conductor–shot electrodes. The length of the test specimen was 100 mm. The distributions of breakdown voltages are shown in Fig. 1. Based on the experimental results, for each batch of conductors the mean values \(\overline{U}\), the root-mean-square deviations \(\sigma\), and the coefficients of variation \(v\) of the breakdown voltages of the turn insulation were calculated; their values are given in Table 1.
Table 1
| Batch No. | Batch characteristics | $\overline{U}$, V | $\sigma$, V | $v$ |
|---|---|---|---|---|
| 1 | Unimpregnated undamaged wire | 756.5 | 70.5 | 0.093 |
| 2 | Unimpregnated damaged wire (longitudinal scratch) | 737.2 | 66.0 | 0.090 |
| 3 | Unimpregnated wire with cuts in the insulation | 731.2 | 61.4 | 0.084 |
| 4 | Impregnated undamaged wire | 1345 | 513 | 0.381 |
| 5 | Impregnated damaged wire (longitudinal scratch) | 621.0 | 121 | 0.195 |
As is evident from Fig. 1, a, b, c and from Table 1, damage to the insulation practically does not lead to a reduction in the characteristics of the distribution of breakdown voltages. This means that even undamaged insulation has numerous air paths from the surface to the conductor, and the breakdown voltage of turn fibrous insulation is the voltage of flashover across the air bridge from the surface of the insulation to the conductor. Since the degree of damage to the insulation in the experiment greatly exceeded the possible degree of damage to the insulation in the manufacturing process of the winding, and under these conditions the breakdown voltage of the insulation did not decrease, in the actual technological process the fibrous insulation of winding wire is not sensitive to technological damage. In this connection, in order to obtain data on the defectiveness of the wire insulation when calculating the reliability of windings, there is no need to test wire removed from the armature slots. The wire can be tested in the as-delivered condition, since it has practically the same insulation characteristics as wire removed from the slots.
Fig. 1, d and the data of Table 1 for the fourth batch of conductors show that many conductors have an insulation breakdown voltage—
Fig. 1. Histograms of the distribution of breakdown voltages of PSDKT wire insulation: a) for undamaged wire; b) for wire with a longitudinal scratch along its entire length; c) for wire with cuts in the insulation; d) for impregnated wire with a longitudinal scratch along its entire length; e) for impregnated undamaged wire.
lying no higher than for the unimpregnated wire. This means that the varnish did not fill the pores of the insulation. The increase in the mean value and variance of the breakdown voltage is associated with the formation of a varnish film on the surface of the insulation. In Fig. 1, d, part of the histogram is shown. Individual outliers of the histogram occur at voltages up to 3.5 kV. The fifth batch of conductors had a longitudinal scratch, that is, the surface film of varnish had been destroyed. As can be seen from Fig. 1, g, the distribution of breakdown voltages in this case has somewhat poorer parameters than was the case for the unimpregnated wire. Consequently, the varnish did not penetrate into the pores of the fibrous insulation and did not make it monolithic. The presence of numerous pores in the layer of fibrous insulation of winding wire both before impregnation and after impregnation requires clarification of the concept of “defectiveness” of the insulation of such wire, for which additional investigations are required.
An analysis of the influence of the winding-work technology on the quality of the turn insulation of PSDKT wires was carried out at the Elektromashina plant, with subsequent testing of the wires in the reliability laboratory of the Department of Electrical Machines of TPI. Wire specimens from one batch were selected in the as-delivered condition, after placing the winding in the slots prior to banding the end parts of the armature, and after banding. The investigations were carried out on armatures of P32 and P52 machines. The wire was removed from the armature slots by a method that excluded additional damage to the wire insulation: on one side, the end part of the winding was cut and bent back. The sections were removed from the slots together with the slot boxes. From the removed wire, specimens 150 mm long were cut from the slot parts of the sections and from the intact end part. The insulation tests were carried out in wire–shot electrodes. The test results are presented in Table 2.
Table 2
| Wire | Characteristic of the distribution of breakdown voltages | Wire as delivered | Wire from the armature before banding | Wire from the armature after banding |
|---|---|---|---|---|
| Wire Ø 1.0 mm | Mean value of breakdown voltage, V | 924,6±2,4 | 921,7±2,4 | 919,6±2,5 |
| Wire Ø 1.0 mm | Root-mean-square deviation of breakdown voltage, V | 47,8±1,69 | 42,1±1,72 | 50,2±1,78 |
| Wire Ø 1.45 mm | Mean value of breakdown voltage, V | 952,5±2,9 | 951,3±2,6 | 945±2,9 |
| Wire Ø 1.45 mm | Root-mean-square deviation of breakdown voltage, V | 59,7±2,06 | 45,8±1,85 | 51,0±2,10 |
The data in Table 2 show that the technological processes of inserting the sections into the slots and bandaging the end parts do not lead to the appearance of insulation defects in PSDKT wire. In contrast to enamel insulation, fibrous insulation is only slightly sensitive to the action of the tool used in laying the winding.
REFERENCES
- Guiding technical materials. Three-phase asynchronous motors for general use with a power from 0.12 to 100 kW. Calculation of the reliability of random-wound windings. OAA.682.022-71.