JPH0432737A - Abnormality detecting device for bearing - Google Patents

Abstract

PURPOSE:To improve the accuracy of abnormality detection by dividing the actual frequency of acoustic emission AE which differs with positions by the logical frequency of AE generation by the positions and thus finding AE generation probability, and comparing and AE generation probability with a single threshold value. CONSTITUTION:When an AE sensor 1 detects AE from the bearing and outputs an AE signal, a BPF 3 removes its noise and after envelope detection 5, a comparator 6 compares the signal with a threshold value and outputs a signal indicating that the signal exceeds the threshold value to a computer 7 when so. The computer 7 calculates the period of the AE signal and totalizes the frequency of the generation of the AE signal in each generation period. The totalized frequencies of generation by generation periods are divided by the logical frequencies of AE generation by the positions corresponding to the generation periods to calculate the probability of AE generation by the positions. The AE generation probability is compared with the single threshold value and when the AE generation probability in a generation period exceeds the threshold value, abnormality of the bearing is decided and a warning is outputted. It can be judged which of the inner ring, rolling body, and outer ring damages from the generation period where the AE generation probability exceeds the threshold value.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a bearing abnormality detection device using an acoustic emitter.

[Conventional technology]

Conventionally, there is a bearing abnormality detection device based on 8E as shown below (Japanese Unexamined Patent Publication No. 63-3 (1983)). This bearing abnormality detection device uses the A and E signals from the AI/ε processor and Comparing means for comparing the values, and based on the output from this comparing means, the AE multiplied signal generation layer I exceeding the threshold value.
I] A frequency 1tl+ calculation means for calculating 1, a totaling means for totaling the number of occurrences for each generation cycle of AC signals based on the output of this period calculating means, and a totaling means for calculating the number of occurrences for each AC signal generation cycle, and a determining means for determining whether or not the threshold value of AE has been exceeded;
If the cumulative value exceeds a threshold value, it is determined that the bearing is abnormal, and the bearing abnormality can be detected even in an environment where AE other than the bearing occurs or where external noise is large.

[Invention or problem to be solved]

However, in the conventional bearing abnormality detection device described in 1.
Since abnormalities are determined by comparing the number of occurrences per AE multiplied signal generation period with a single threshold value, there is a problem that detection accuracy is low for the following reason. For example, during one rotation of the bearing, if there is an abnormality in the inner ring, the E signal will be generated only once when the threshold is exceeded, but if there is an abnormality in the outer ring, the AE signal will be generated when the threshold is exceeded. This occurs approximately (number of rolling elements/2) times. Therefore, the threshold value to be compared with the number of AE occurrences is different for each part such as the inner ring and the outer ring. However, since the conventional bearing abnormality detection device described above compares the number of AE occurrences with a single threshold value for all parts, the accuracy of abnormality detection is low. Therefore, the purpose of this invention is to
A bearing abnormality detection device that can detect bearing abnormalities with high accuracy by dividing the number of occurrences by the theoretical number of AE occurrences for each part to find the probability of AE occurrence, and comparing this AE occurrence probability with a single threshold. Our goal is to provide the following.

[Means to solve the problem]

In order to achieve the above object, the bearing abnormality detection device of the present invention includes an AE sensor that detects acoustic emission from the bearing and outputs an AE multiplied signal, and an AE multiplied signal threshold from the AE sensor. and a comparison means to compare.”
a period calculating means for receiving a signal indicating that the E signal exceeds the threshold value from the comparing means in item 1, and calculating an AE multiplied signal generation period exceeding the threshold value; a totalizing means for totalizing the number of occurrences for each occurrence period calculated by the period calculation means, and dividing the number of occurrences for each occurrence period calculated by the aggregation means by the theoretical number of occurrences for each site corresponding to each occurrence period; and an AE occurrence probability calculation means for calculating the AE occurrence probability for each part, and determining whether or not the AE occurrence probability outputted from the AE occurrence probability calculation means exceeds a threshold value to determine whether there is an abnormality in the bearing. The present invention is characterized in that it is equipped with a discriminating means.

[Effect]

AE from the bearing is detected by the AE sensor, and a signal is output. This E signal is compared with a threshold value by a comparing means, and when the AE multiplied signal exceeds the threshold value, a signal representing this is output. In response to the signal from the comparing means, the period calculating means calculates the AE multiplied signal period. The aggregation means aggregates the number of occurrences of the AE multiplied signal for each generation cycle calculated by the cycle calculation means. The above AE
The occurrence probability calculation means divides the number of occurrences for each occurrence period, which is totaled by the aggregation means, by the theoretical number of occurrences for each region corresponding to the occurrence period, thereby calculating the probability of AE occurrence for each region. The above-mentioned discrimination means is an AE output from the AE occurrence probability calculation means.
The occurrence probability is compared with a single threshold value, and when the AE occurrence probability in the occurrence period exceeds the threshold value, it is determined that the bearing is abnormal. In this way, since the probability of AE occurrence is compared with the threshold,
There is no need to set different threshold values for each part, and bearings can be diagnosed with high accuracy using a single threshold value. - It can be determined whether the inner ring 1 rolling element or the outer ring is damaged, based on the occurrence cycle in which the AE occurrence probability exceeds the threshold value.

【Example】

Hereinafter, the present invention will be explained in detail with reference to illustrated embodiments. In Fig. 1, 1 detects AE of bearings etc. Δr>
It is a sensor. After the AE multiplied signal representing the magnitude of 8E output from this AE sensor 1 is amplified by the preamplifier 2, the AE multiplied signal in the band from, for example, 100KI-1z to 500KIIz is passed through the bandpass filter 3, and noise is removed. be done. The AE multiplied signal from which noise has been removed by the band pass filter 3 is further amplified by the main amplifier 4, input to the envelope detection circuit 5, and subjected to envelope detection. The AE multiplied signal has a waveform as shown in FIG. 2(a), for example, and a spectrum as shown in FIG. 2(b). The waveform after envelope detection by the envelope detection circuit 5 is as shown in FIG. 2(C). After envelope detection, the AE multiplied signal is compared with a threshold in the AE multiplied signal comparator 6, and when the AE multiplied signal exceeds the threshold, a pulse as shown in FIG. 2(d) is outputted to the computer 7. Further, the number of rotations of the bearing per unit time is inputted to the computer 7 from the rotation sensor 8 . In the combination coater 7, processing is performed according to a flowchart as shown in FIG. First, initial settings are made in step S1 in FIG. 3, and the process proceeds to step S2, where it is determined whether a predetermined period of time has elapsed. Here, if it is determined that the specified time has not elapsed, the process proceeds to step S3, and it is determined whether or not AE has occurred depending on whether or not a pulse has been received from the comparator 6. If it is determined that no AE has occurred, the process returns to step S2, and if it is determined that an AE has occurred, the process proceeds to step S4, where the time A (1
) is stored in the memory, and the process returns to step S2. The time A (1) at which this AE occurred is shown in FIG. ”-Step S2. S3. After repeating S4, step S
If it is determined in step 2 that the specified time has elapsed, the process proceeds to step S5, and as shown in FIG. 4.5, the AE occurrence times A(+)Δ(2), . . . , A( 4)
Occurrence period 3 representing the time between), 13(2),...
13(6) is calculated. This generation cycle B(1) is calculated by the routine shown in FIG. First, in step S21, I-1, N=
I, is initialized as = I. Then,
Proceeding to step S22, the occurrence period B(1) is fA(2
)−A(1)), and then step S2
3, it is determined whether N has become (number of occurrences - 1). Now, as shown in Figure 5, the number of occurrences is 4 and I=
Since it is 1, it is determined whether it is N or 3. Now, since N is 1, the process advances to step S24 and K=2. N=2
, the process proceeds to step S22 and the occurrence period B(2)=
(Δ(3)-A(1)), and further step S2
3. After S24 and S22, the generation period B(3)=(A(
4) Find -A(I)). Next, since N=3 in step 823, the process proceeds to step S25, and
-(number of occurrences - 1), that is, whether or not the occurrence period B(6) has already been found at I=3. Since I=1 now, proceed to step 826 and set I=2.
With N=1, the process returns to step S22 and the occurrence period B
(4) is obtained, and further, step S23. S2/I,S
22, find the generation period B(5). Furthermore, step S23. After S25, 526 and S22, the generation cycle B (
6), and step S23. After S25, the calculation of the occurrence period shown in FIG. 7 is completed. Then, the process returns to step S6. In step S6, the number of revolutions per unit time of the bearing received from the rotation sensor 8 during a specified period of time is converted into a reference number of revolutions, and the period calculated in step S5 is corrected to the period to reach the reference number of revolutions. That is, the circumference M calculated in step S5 is multiplied by the number of revolutions per unit time of the bearing detected by the rotation sensor 8, and divided by the reference number of revolutions.Next, the process proceeds to step S7, and each AE Aggregation is performed for each double signal generation period. That is, as shown in FIG. Calculate the probability of E occurrence by dividing the number of E occurrences by the theoretical number of AE occurrences for each cycle (by location).Here, the theoretical number of Δ■C occurrences is for each location of the inner ring, outer ring, and rolling element. , is the number of AEs generated from each part per rotation of the bearing, assuming that there is damage in one part.Next, proceed to step S9,
As shown in Figure 6, it is determined whether the A2 occurrence probability exceeds a single threshold value for each occurrence cycle, and if the A2 occurrence probability exceeds the threshold value, bearing damage is determined. After making a determination, the process proceeds to step SIO and outputs an alarm indicating an abnormality in the bearing. In this way, the threshold value is compared with the A2 probability of occurrence, which is calculated by dividing the number of AE occurrences that vary by region by the theoretical number of AE occurrences, so damage to each region can be detected with high accuracy using a single threshold value. Ru. The damaged location of the bearing can be identified based on the occurrence cycle in which the AC occurrence probability exceeds the threshold value. Zunawaji,
If the above-mentioned bearing is a roller bearing with the inner ring as a rotating ring, Fi is the frequency at which the rollers pass through a certain part of the inner ring, Fr is the rotational frequency of the shaft, and F is the frequency at which the rolling elements pass at a certain part of the outer ring. , when the rotation layer wave number of the roller is Fb and the rotation frequency of the cage is Pc, if there is an abnormality in the inner ring, I/Fi, l/Pr if there is an abnormality in the outer ring, l/F'. If there is an abnormality in the roller, AE will occur at the I/Fb and I/Fc cycles. The cycle of AE occurrence differs depending on the damage to the inner ring, outer ring, and rollers. I want to, A.
It is possible to determine whether peeling has occurred in the inner ring, outer ring, or roller based on the occurrence cycle in which the E occurrence probability exceeds a threshold value. If it is determined in step S9 that the AE occurrence probability does not exceed the threshold in any occurrence period, the process returns to step S2. In this example, after removing noise with a bandpass filter having a band of IOOK, l-1z to 500KHz,
Detects an AE multiplied signal with an amplitude above a certain level, calculates the number of occurrences per period of this AE multiplied signal, and calculates the AE occurrence probability. Since it is determined that there is an abnormality in the bearing, it is possible to accurately detect abnormalities such as peeling that occur in the bearing, and also to identify the peeled part of the bearing. Additionally, changes in the bearing rotational speed will affect the AE occurrence cycle, but since the rotational speed from the rotation sensor is detected and the occurrence cycle is corrected, the effect of fluctuations in the bearing rotational speed can be reduced. It is possible to accurately identify the damaged location of the bearing without having to worry about damage. In addition, since bearing abnormalities are determined based on the E occurrence probability of the AE multiplied signal that occurs with a specific period, it is possible to detect abnormalities in bearings in environments where AE other than bearings occurs, such as during metal-in of a rolling mill. It is possible to accurately detect abnormalities such as peeling, and to identify the peeled part of the bearing. In the embodiments described above, the threshold value for comparison with the AE amplitude is set to a constant value, but the threshold value may be varied depending on the magnitude of external noise or the like. Furthermore, in the embodiment described in -1- above, the AE generation cycle is corrected after the specified time has elapsed, but the AE generation cycle may be corrected each time an AE occurs. moreover,
If the rotation speed is constant or a change in rotation speed is expected,
Instead of using the rotation sensor 8, the rotation speed and predicted rotation speed may be manually entered into the computer 7 to correct the AE occurrence cycle. If the rotation speed is constant and it is not necessary to determine the damaged portion of the bearing, it is possible to not correct the AE occurrence cycle.

【Effect of the invention】

As is clear from the above, the M receiver abnormality detection device of the present invention includes a comparison means for comparing threshold values of the AE multiplied signal from the AE sensor, and a period for calculating the generation period of the AE multiplied signal that exceeds the threshold. a calculation means, a totalization means for totalizing the number of occurrences for each AE multiplication signal generation cycle, and an AE occurrence probability calculation means for dividing the number of occurrences of AE by the theoretical number of AE occurrences. , a determination means for determining whether or not the AE occurrence probability exceeds a threshold value is provided, and an abnormality in the bearing is determined by comparing the AE occurrence probability for each occurrence cycle with a single threshold value. Bearing abnormalities can be detected with high precision and the location where they occur using a single threshold value.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the bearing abnormality detection device of the present invention.
FIGS. 2(a), (b), (c), (d), and (e) are diagrams showing waveforms in each part of this embodiment, and FIG. 3 is a graph of the above embodiment [J-chart, Figure 4 is a flowchart showing the subroutine for calculating the AE occurrence cycle, Figure 5 is a diagram showing the relationship between the occurrence time and the occurrence cycle, and Figure 6 is a diagram showing the relationship between the occurrence cycle, the AE occurrence probability, and the threshold value. be. ATε sensor 3 Band pass filter 5... Envelope detection circuit, 6... Comparator, 7... Combi coater,
8... Rotation sensor. Patent applicant: Koten Seiko Co., Ltd. and one other representative
Person Patent attorney Aoyama Ao and one other person Figure 2 (a) (b) Busy) +d) (e) Figure 3

Claims (1)

[Claims] (1) an AE sensor that detects acoustic emissions from a bearing and outputs an AE signal; a comparison means that compares the AE signal from the AE sensor with a threshold value; A that exceeds the above threshold upon receiving a signal indicating that the threshold has been exceeded.
a period calculation means for calculating the generation period of the E signal; a totaling means for totaling the number of occurrences for each generation period calculated by the period calculation means; Divide by the theoretical number of occurrences for each site corresponding to each generation cycle,
AE occurrence probability calculation means for calculating the AE occurrence probability for each part; and determination means for determining whether or not the AE occurrence probability outputted from the AE occurrence probability calculation means exceeds a threshold value to determine whether there is an abnormality in the bearing. A bearing abnormality detection device characterized by comprising: