JPH0726941B2 - Bearing life prediction method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a bearing life prediction method using acoustic emission (AE).

<Prior Art> Conventionally, a bearing abnormality detection method using an acoustic emission signal (AE signal) has been used for sliding bearings and has been as follows. For example, the abnormality of the bearing was detected by detecting the AE signal emitted by the damage of the metal surface caused by the mutual contact between the sleeve and the bush when the oil film of the sliding bearing was broken.

<Problems to be Solved by the Invention> However, the above-mentioned conventional method for detecting abnormality of a bearing detects a state in which a metal surface of a sleeve or a bush of a sliding bearing is damaged and is already damaged. There was a problem that I could not grab.

Therefore, in the past, bearings for aircraft that affect human life, bearings for vehicles, or 24-hour continuous operation have been required,
For generator bearings, water supply facility bearings, etc., where failures are absolutely unacceptable, the bearings are replaced at an extremely high safety factor, and even if they are fully usable, many are discarded. was there. On the other hand, in some cases, the bearing was destroyed and suffered a great deal of damage even under a high safety factor.

Therefore, an object of the present invention is to predict the life of a bearing by detecting a change in the internal structure of the bearing, which is a precursor to destruction, by acoustic emission.

<Means for solving the problem> An AE signal having a frequency spectrum as shown in Fig. 2 is detected from the bearing. However, this Figure 2 is 100KHz
Since the following is greatly affected by mechanical vibration, it is removed by a filter. This AE signal is mainly generated by the plastic deformation of the transfer surface, the structural change inside the bearing member, and the slip of the rolling elements due to the structure of the bearing.

The present inventor has found that the spectrum of the AE signal due to the above-mentioned plastic deformation or slippage that causes fracture is as shown in FIGS. These spectra are characterized by less than 200 KHz, and the effective value of the signal is also much smaller than the effective value of the entire AE signal detected from the bearing, being less than 1/10. Therefore, it is possible to detect the plastic deformation and the slip of the rolling elements by detecting a spectrum similar to the spectrum shown in FIGS. Therefore, it can be seen that, of the spectrum signals generated in FIG. 2, 200 KHz or higher is the spectrum due to internal tissue change. Fifth
The figure shows the microstructural changes that occurred inside the bearing member.

Next, the AE detected from the bearing with cracks inside
The spectrum of the signal is shown in FIG. 6th compared to FIG.
The spectrum in the figure shows an increase in power value from 150 to 300 KHz. It was found that the AE signal at 150 to 300 KHz was correlated with the occurrence of microcracks. FIG. 7 shows the internal crack of the bearing which emits the AE signal shown in FIG.

As described above, the present inventor has discovered that the spectrum of the AE signal generated by the structural change inside the bearing member, which is a precursor of the destruction of the rolling bearing, is characterized by 200 KHz or more. In addition, the spectrum of the AE signal due to the occurrence of cracks inside the bearing, which is a precursor of more advanced destruction, is 150 to 300.
It was discovered that it has characteristics in KHz.

The present invention is based on the above findings, in order to detect the precursor of the fracture, not to the already fractured state, but to detect the plastic deformation of the rolling bearing, mainly the plastic deformation of the transfer surface, the sliding of the rolling element, and the microstructural change inside the bearing member. In the bearing life prediction method that predicts the life of the bearing from the acoustic emission (AE) generated from the bearing, it is extracted by a bandpass filter that passes the output of the band from 150 KHz to 300 KHz among the output from the AE sensor. It is characterized by comparing the AE signal spectrum in the above frequency band with a reference value, and determining the presence or absence of cracks due to the microstructural change inside the bearing member by the presence or absence of the AE signal spectrum between 200 KHz and 300 KHz. There is.

<Example> Hereinafter, the present invention will be described in detail with reference to illustrated examples.

In FIG. 1, 1 is a sensor attached to a bearing to detect acoustic emission from the bearing, 2 is a preamplifier, 3 is a hand-pass filter that passes an AE signal in the band of 150 KHz to 300 KHz, 4 is a main amplifier, and 5 is Envelope detection circuit, 6 is an A / D converter, 7 is the AE signal input from the A / D converter 6 and the reference value are compared, and when the above signal exceeds the reference value, as a sign of destruction It is a computer as an arithmetic unit for detecting and instructing to issue an alarm.

With the above configuration, the AE signal detected by the sensor 1 is input to the bandpass filter 3 via the preamplifier 2 and only the frequency component in the range of 150 to 300 KHz is extracted. This output is further amplified by the main amplifier 4,
After envelope detection is performed by the envelope detection circuit 5, A / D conversion is performed by the A / D converter 6 and the result is taken into the computer 7. In the computer, the AE signal spectrum in the band of 150 to 300 KHz is compared with the reference value, and 200 K higher than the reference value is exceeded.
The alarm is issued when the AE signal spectrum between Hz and 300 KHz is detected.

Thus, according to the bearing life prediction method of this embodiment,
Of the AE signal generated from the bearing, frequency components (150KHz to 300K) that correlate with internal microstructural changes and internal cracks.
It is possible to easily and accurately predict the occurrence of cracks in the bearing by discriminating the AE signal with (Hz). Furthermore, since the presence / absence of a crack due to the microstructure change inside the bearing member is judged by the presence / absence of the AE signal spectrum between 200 KHz and 300 KHz, it is possible to selectively judge only the crack due to the microstructure change inside the bearing. You can

In the above-described embodiment, the computer 7 compares the signal from the A / D converter 6 with the reference value to discriminate the precursor of destruction.
The precursor of destruction may be detected by comprehensively determining the number of AE signals generated, the waveform shape, the amplitude, the generation state, and the like.

<Effects of the Invention> As is apparent from the above, the bearing life prediction method of the present invention is such that the frequency band extracted by the bandpass filter that passes the output in the band between 150 KHz and 300 KHz among the outputs from the AE sensor. By comparing the AE signal spectrum of No. 2 with the reference value, the presence or absence of cracks due to the microstructural change inside the bearing member is determined by the presence or absence of the AE signal spectrum between 200 KHz and 300 KHz.

Therefore, according to the present invention, it is possible to selectively extract only the AE signal caused by the plastic deformation of the transfer surface of the rolling bearing, the slip of the rolling element, and the change in the structure of the bearing member. Therefore, since only the signals necessary for predicting the life of the bearing are extracted in advance, the life of the bearing can be predicted easily and accurately.

Further, the presence or absence of an AE signal spectrum between 200 KHz and 300 KHz is used to determine the presence or absence of cracks due to the structural change inside the bearing member, so that only the cracks due to the structural change inside the bearing member are selectively determined. be able to.

Therefore, according to the present invention, it is possible to accurately determine the timing of bearing replacement before destruction according to each application, maintain the operating accuracy of the device, secure the safety of human life, etc. It is also possible to eliminate the inconvenience of replacing a bearing that is fully usable.

[Brief description of drawings]

FIG. 1 is a block diagram of an apparatus for executing a bearing life prediction method according to an embodiment of the present invention, FIGS. 2, 3, 4, and 6 are diagrams showing a power spectrum of an AE signal, FIG. 5 and FIG. It is a figure which shows the crystal structure of a bearing member. 1 ... Sensor, 2 ... Preamplifier, 3 ... Bandpass filter, 4 ... Main amplifier, 5 ... Envelope detection circuit,
6 ... A / D converter, 7 ... computer.

Claims (1)

[Claims] 1. A bearing in which the life of the rolling bearing is predicted from the acoustic emission (AE) generated from the bearing mainly due to plastic deformation of the transfer surface, sliding of the rolling elements, and structural change inside the bearing member. In the life prediction method, compare the AE signal spectrum of the above frequency band extracted by the band pass filter that passes the output of the band between 150 KHz and 300 KHz among the outputs from the AE sensor with the reference value, and from 200 KHz to 300 KHz. A method for predicting bearing life, characterized in that the presence or absence of an AE signal spectrum between the two is used to determine the presence or absence of cracks due to structural changes inside the bearing member.