JPH0277664A - Method for diagnosing insulation of winding of electric machine - Google Patents

Abstract

PURPOSE:To easily diagnose an extent of deterioration of insulation nondestructively by detecting change in mechanical qualities of winding insulation due to thermal deterioration and mechanical deterioration of the winding insulation by an acoustic emission (AE) method. CONSTITUTION:Four windings 2 of a DC machine whose insulation is to be diagnosed are located combined with a field core 3 at intervals of 90 deg. on the inner of a frame 1, and wiring is made to permit power to be supplied. AE generated at the time of change in temperature of the winding 2 is detected by an AE sensor 4 provided on the outer of the frame 1. A parameter obtained by detecting AE, which the insulation layer of said winding 2 is subject to, due to change in temperature with the AE sensor 4 at the time of newly producing the winding 2 is used as reference, and a parameter obtained by detecting AE generated due to change in temperature of the winding 2 to be inspected is compared with the above parameter to diagnose the extent of deterioration in insulation.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an insulation diagnosis method for measuring the degree of insulation deterioration of windings of an electric machine using an acoustic emission (hereinafter referred to as "E") method.

[Conventional technology]

In general, a major factor that affects the reliability and lifespan of electric machines such as generators and motors is insulation deterioration of windings. Furthermore, accurate evaluation of the degree of deterioration of the insulation is desired when renewing the insulation at the end of the life of the equipment.

Non-destructive insulation deterioration diagnostic tests include electrical, chemical,
Various mechanical methods have been studied, but electrical methods are mainly used to investigate the characteristics of leakage current and partial discharge when DC or AC voltage is applied to the windings. This is a method for determining the properties of an insulator.

In addition, as a method for mechanically diagnosing insulation deterioration, hammering sounds are used, but this is done by locally applying an external force from the outside when the temperature of the insulating layer of the winding is static and the temperature does not change. In addition, this method attempted to evaluate the insulating layer based on the reaction.

[Problem that the invention sought to solve]

However, in the case of electrical tests, the degree of insulation deterioration was diagnosed by detecting defects such as cracks, holes, and voids within the insulating layer, but the degree of mechanical elasticity of the insulating layer was not detected. could not.

In addition, when investigating mechanical elasticity, it was not possible to quantify it using the hammering sound, and individual differences occurred, so this was not the preferred method.

In addition, these diagnostic methods are performed under a static state in which the temperature of the insulating layer of the winding does not change, so it is not possible to perform the diagnosis under a dynamic state in which the temperature change of the winding is the same as or close to the actual operating state. There was a desire to develop a diagnostic method for

The present invention is a method devised in view of the above-mentioned points, and its purpose is to use the AB method to detect changes in the mechanical properties of winding insulation due to thermal deterioration and mechanical deterioration of winding insulation. The object of the present invention is to provide a method for diagnosing the winding insulation of an electric machine, which can easily and non-destructively diagnose the degree of insulation deterioration.

[Means for solving problems]

In other words, the means to achieve this purpose is based on the basic application of Japanese Patent Application No. 161908/1986, which was filed earlier. (Mechanical stress caused by the difference in the expansion coefficient of the material caused by the temperature difference between the Therefore, insulation diagnosis is carried out by measuring and analyzing the human E that occurs under such temperature change conditions ξζ. For the diagnosis, it is necessary to compare with the AE data of the same insulation system when the insulation was newly manufactured.

That is, AB parameters (for example, the number of occurrence events) that occur due to temperature changes when new insulation is manufactured.

This is a method of diagnosing the degree of insulation deterioration by comparing the parameters of AI generated due to temperature changes in the winding to be inspected with the parameters at the time of new manufacture, using the accumulated energy (accumulated energy, etc.) as a reference value.

Next, the claim number filed this time (in terms of claim number, diagnosis 1
This means that a group of measured and analyzed AI data is quantified to determine what kind of distribution shape the AI data has based on a specific parameter of AH, such as duration. The method of quantifying is to divide the value of a specific parameter, for example, the duration, into multiple bands and calculate the ratio of the number of AI data being idle between multiple (usually two) specific bands. .

This is a method of diagnosing the degree of insulation deterioration based on the numerical value of the ratio obtained as described above.

[For production]

The effect is that when the electric machine is operated, the temperature of the insulating layer is heated by the Joule heat of the flowing current and rises, and when the electric machine is stopped, it is cooled and the temperature falls.

The insulating layer iζ between the conductor that is simultaneously generating heat and the cold core or air has a temperature gradient that changes over time. For this reason, the insulating layer is subjected to thermal strain, and to release this strain, crazes, cracks, and cracks occur within the insulating layer, as well as peeling and friction between the insulating layer and other interfaces, and at the same time, AE occurs. Occur.

When insulation deterioration progresses in this way, the appearance of these phenomena, which release thermal strain within the insulation layer, differs depending on the insulation system, but it is possible to Insulating systems in which resin is vacuum-impregnated can be considered as follows.

(υ Claim (1) states that when the insulation system is first manufactured, the core, insulation layer, and conductor are solidified by resin and are completely integrated. At the beginning of deterioration, thermal strain occurs in the insulation layer. When stress is applied, crazes, cracks, and cracks occur in the insulating layer.However, since the insulating layer is still new and the resin layer has elasticity to absorb strain, only a certain limit of AE can be measured (the first step in the insulation deterioration process). step).

As insulation deterioration progresses a little further, the resin layer heats up.

After repeated cooling, it hardens and becomes an elastic material that absorbs strain, which causes crazes, cracks, and cuts in the film, as well as peeling and friction, and the occurrence of AE also reaches its peak (second stage).
As the insulation deteriorates further, the thermal strain that occurs due to the large number of crazes, cracks, and cracks that have already occurred in the insulation layer decreases, and the AE that occurs with it also decreases (Step 3). Stage) When insulation deterioration reaches its final stage, the thermal strain within the insulation layer will further decrease, and when the number of AE events that occur decreases to about 30% of the initial level, the dielectric breakdown voltage of the insulation layer will decrease to about 20% of that when it was new. The number decreases (4th stage).

(2!J Claim No. 2!J, when the insulation system is first manufactured, the iron core, insulation layer, and conductor are solidified with resin and are completely integrated. At the beginning of deterioration, the insulation layer may have crazes or Crazes and cracks are likely to occur when thermal strain is applied because no cracks or cracks occur.If you analyze the AE measured at that time, for example, by looking at the distribution of duration, you will find that those with short durations are overwhelmingly There are many cases, but there are some that last for a long time.This suggests that although most of the cases are small radar or cracks, there are a few large cases.

As insulation deterioration progresses, cracks and cracks have already occurred in areas where thermal strain is likely to occur within the insulation layer.
Thermal strain also decreases, and the number of AEs that occur decreases, but the occurrence of AEs that last a long time, that is, the occurrence of large AEs, decreases compared to those that last a short time. Thermal strain within the insulation layer is further reduced, the mechanical elasticity of the insulating layer is almost eliminated, and among the AE events that occur, the occurrence of long-duration ripples is further reduced. When the ratio of the number of AE events of short duration to those of long duration decreases to a certain value, the dielectric breakdown voltage of the winding decreases to about 20% of that of a new one.

Hereinafter, one embodiment of the insulation diagnosis method of the present invention will be described in detail based on rotation.

, [Example of implementation] Figures 1 (a) and (b) show an example of Akira Hongo's insulation diagnosis method, ``showing an assembly in which the DC machine field winding of a converting machine is assembled into an iron core and a frame.'' Figure 1 (b) is a half-view diagram showing the human E measurement method.
1) is a side view of the assembled product.

In Fig. 1 (, 10, (b), the winding 2 of the DC machine for which insulation diagnosis is performed is placed between the field cores 3 and 3 at 90° intervals within the frame 1.
Four of them are arranged in combination with each other, and they are wired so that they can be energized.

The AE that occurs when the winding 2 is in a state of temperature change is detected by the AE sensor 4.
The AE sensor 4 is provided with a plane on the outer periphery of the frame 1, and the overwriting surface coated with grease is mechanically fixed so as to be in close contact therewith.

Note that 5 is a preamplifier, 6 is an AE analyzer, and 7 is a printer.

Next, a method for diagnosing insulation deterioration of windings set in such conditions will be explained with reference to Figures 2 (a) and (b) to Figure 4.The detected AE signal , a preamplifier 5, and is analyzed by an AI analyzer 6, and the analysis results are displayed on the AH analyzer 6 as well as output to a printer 7#ζ. AH used
The sensor 4 has a resonant frequency of 150 k) h iC, and the gain of the preamplifier 5 is 40 dB.

The main amplifier gain in AH analyzer 6 is 20dB.

The threshold value was set to 0.1v for measurement. In addition, as an insulation system for the field winding, we have prepared an insulation system that uses ground insulation iζ film-wound insulation and vacuum-impregnated solvent-free resin. Apply a slightly lower constant current and heat for 5 hours.
A heat cycle was performed with a cooling time of 1 hour.

Figure 2 (- is a characteristic diagram of the resistance temperature of the winding due to the heat cycle and the elapsed time at a certain number of heat cycles, and Figure 2 (b) is the number of AI events that occurred in the winding at this time and the elapsed time. FIG.

From the second open can), it can be seen that a large number of AEs occur when the temperature rises rapidly at the beginning of the heat cycle elapsed time.

Next, it will be explained in claim (1).

In Figure 3, the vertical axis shows the ratio of the number of AE events that occur in one heat cycle shown in Figure 2 (b) to the number of AI events that occur in the heat cycle when newly manufactured, and the horizontal axis indicates the number of heat cycles, and is a characteristic diagram showing how the number of AE events changes as the heat cycle progresses, and the numbers 1 to 4 at the top of the characteristic diagram indicate [effect]. It shows the process steps of insulation deterioration as explained in section.

From Figure 3, it can be seen that the impregnated resin retains its elasticity from the beginning of the heat cycle to around 100 cycles, which is the first stage of insulation deterioration.
・The second stage is followed by the third and fourth stages in which the thermal strain is absorbed by cracks, etc., and the number of occurrence events gradually decreases.

Next, in Figure 4, the vertical axis in Figure 3 is expressed as the ratio of the number of AE events that occur in one heat cycle (6 hours); FIG. 3 is a characteristic diagram showing the ratio of the number of AE events.

After that, the four field windings 2 that were subjected to the heat cycle test showed a voltage of 5kV after 600 to 700 cycles due to insulation deterioration.
Dielectric breakdown occurred due to the withstand voltage.

From this, the ratio of the number of AB events is 3o in Figure 3.
*, Fig. 4 shows that there is a possibility of dielectric breakdown when the concentration decreases to about 2O4. In this embodiment, the number of AE events is used as a parameter for diagnosing the degree of insulation deterioration, but instead of this, the cumulative value of AI energy and the cumulative value of counts are used.

Diagnosis can also be made by the cumulative value of duration. Furthermore, even in the same insulation system, the number of AP events that occur varies depending on the shape of the windings, so it is necessary to measure the characteristics when newly manufactured each time and perform diagnosis by comparing with the characteristics.

Further, to explain claim (2), FIG. This is a characteristic diagram in which the horizontal axis is the duration (logarithm display) and the vertical axis is the occurrence rate of AH events as a percentage for the AE data of the first 10% of events, that is, the part where many AEs occur during heating. be.

This shows what kind of duration distribution the generated AI has, and the reason why the duration is displayed logarithmically is to emphasize the number of A1 occurrences on the long side compared to those on the short side. It is from. Figure 5 (supplemented) is at the start of the heat cycle test, Figure 5 (b) is around 200 cycles, Figure 5 (
ej shows the characteristics around 600 cycles @ Figure 6 is the ratio B of the AB event occurrence ratio between the double diagonal line area A going up to the right and the single diagonal line area B going down to the right in Figure 5.
This figure shows how /A changes depending on the number of heat cycles. According to this, at the start of the heat cycle test, it is around 1.4, and as the heat cycle test progresses, it monotonically decreases, and around 200 cycles, it is around 1.1,60.
Around 0 cycles, the value is between 0.6 and 0.7.

The four field windings 2 that were subjected to the heat cycle test as described above had a temperature of 600 mm due to thermal deterioration due to the heat cycle.
Dielectric breakdown was caused by a 5 kV withstand voltage check during ~700 cycles. From this, it was found that there is a possibility of dielectric breakdown when the ratio B/A of the AI event occurrence rate decreases to around 0.6. In this embodiment, the duration of the AE event was used as a parameter for diagnosing the degree of insulation deterioration, but instead of this, energy, ring down count, etc. may also be used for diagnosis.

The characteristic diagrams in Figure 5 (a), (b), and (c) are obtained by analyzing the first 10% of AI data during one heat cycle, but these characteristics are based on the analysis of all AE data during one heat cycle. The characteristics are almost the same. Therefore, the time required for one diagnosis may be significantly less than the six hours required for one heat cycle, and 15 to 20 minutes is sufficient.

〔Effect of the invention〕

As explained above, according to the present invention, although conventionally it was not possible to quantitatively detect changes in the mechanical properties of an insulating layer, the AE method is used to detect AE generated from thermal strain in an insulating layer. Through measurement, it has become possible to easily and quantitatively diagnose the degree of deterioration of winding insulation in a non-destructive manner. In addition, in claim (2), the present insulation diagnosis method is not a method of diagnosing by calculating the AE data of the winding to be diagnosed when it is newly manufactured and the AE data at the time of diagnosis. There is no need for data when the new product is manufactured. Therefore, AE data measurement when newly manufacturing the winding can be omitted.

[Brief explanation of the drawing]

1(a) and 1(b) show an embodiment of the diagnostic method of the present invention, FIG. 1(a) is a plan view thereof and an explanatory diagram showing the AB measurement method, and FIG. 1(b) is a side view. Figure 2(a) is a characteristic diagram showing the relationship between winding resistance method temperature and elapsed time in a certain heat cycle, and Figure 2(b) is a characteristic diagram showing the relationship between winding resistance method temperature and elapsed time in a certain heat cycle. Figure 3 is a characteristic diagram showing the relationship between the number of AE events occurring in a line and elapsed time, Figure 3 is a characteristic diagram showing the relationship between the ratio of AE events occurring in one heat cycle and the number of heat cycles, and Figure 4 is a characteristic diagram showing the relationship between the number of AE events occurring in one heat cycle and the number of heat cycles. Characteristic diagrams showing the relationship between the ratio of AE events that occur during the first 5 minutes of a heat cycle and the number of heat cycles, Figures 5 (a), (b), and (C) show the relationship between the AE events that occur during the heat cycle and the number of heat cycles. Figure 5 (a) shows the distribution at the start of the heat cycle test, Figure 5 (b) shows the distribution around 200 cycles,
Figure 5 (e) is a characteristic diagram around 600 cycles, and Figure 6 shows how the ratio B/A of the human E event occurrence rate in parts A and B in Figure 3 changes depending on the number of heat cycles. FIG. 1... Frame, 2... Field winding, 3...
...Field iron core, 4...AE sensor, 5...
...Preamplifier, 6...AE analyzer, 7
...Printer.

Claims (2)

[Claims] (1) In a method for diagnosing insulation deterioration of an electromechanical winding, when the winding is newly manufactured, acoustic damage occurs due to thermal strain that the insulating layer of the winding undergoes due to temperature changes.
The parameters obtained by detecting emissions with an acoustic emission sensor are used as reference values, and the parameters obtained by detecting acoustic emissions generated due to temperature changes in the windings subject to insulation diagnosis are compared with the reference values. A method for diagnosing insulation of electrical machine windings, characterized by diagnosing the degree of insulation deterioration by (2) When the winding is newly manufactured, the parameters obtained by detecting the acoustic emissions generated by the thermal strain that the insulating layer of the winding undergoes due to temperature changes with an acoustic emission sensor are used as reference values, and the insulation A method for detecting acoustic emissions caused by temperature changes in the windings to be diagnosed and comparing the obtained parameters with the reference values. Detected and measured by an acoustic emission sensor, examine how the number of acoustic emission occurrence events is distributed according to specific parameters of acoustic emission, quantify the distribution shape, and insulate it with that value. A method for diagnosing insulation deterioration of a winding of an electric machine according to claim 1, characterized in that the degree of deterioration is diagnosed.