JP4787904B2 - Rolling bearing remaining life diagnosis method - Google Patents

Description

  The present invention relates to a pump that is an auxiliary machine used in chemical plants, steelworks, power plants, etc., a rolling bearing for a fan, and a rolling bearing that estimates the remaining life of a rolling bearing used in a motor that drives these devices. The present invention relates to a remaining life diagnosis method of a bearing.

  In the rolling bearings used in pumps and fan rolling bearings, which are auxiliary machines used in chemical plants, steelworks and power plants, and motors that drive these devices, the load load is 5% or less of the rated load. It is very small and does not cause metal fatigue under normal use. Therefore, the life of these rolling bearings is the surface roughness of the raceway surface of the rolling bearing due to the “peeling due to stress concentration” at the rising part of the indentation generated due to the mixing of dust or the oil film breakage of grease due to the mixing of water. There are two types of “increase in vibration” due to an increase in.

Various means have been proposed as a method of diagnosing the remaining life of the rolling bearing. For example, as shown in “Bearing Diagnosis Method” of Patent Document 1, the vibration of a bearing is measured using an acceleration sensor signal, and an alarm is issued when the bearing vibration value exceeds an allowable value. Acoustic emission (Acoustic) Emission) method and the like have been proposed. In addition, there is a method of estimating the cause of the failure by analyzing the frequency of the bearing vibration. There is also a method for predicting the service life by predicting the increasing tendency of the bearing vibration value.
JP-A-8-159151

  The method of predicting an increasing tendency of a bearing vibration value using an acceleration sensor signal is the most frequently used prediction method. This is a method of predicting the increasing tendency of the acceleration vibration of the bearing by a straight line, a quadratic curve, or an exponential curve, and predicting the remaining life of the bearing by the remaining time until reaching a preset allowable vibration value.

For example, as shown in FIG. 12, an effective value of 0 to 10 kHz is calculated for the acceleration vibration waveform, and the determination is made by measuring two kinds of threshold values, an absolute value and a relative value, and exceeding this threshold value. If not, it is determined that the diagnosed rolling bearing is “normal”.
On the other hand, when the threshold value is exceeded, it is determined that the diagnostic rolling bearing is “abnormal”, and the frequency spectrum of the vibration waveform is calculated. Extract n-times components of the rotational speed such as 1N, 2N, 3N, and mN.
Alternatively, when it is determined that the diagnostic rolling bearing is “abnormal”, the envelope processing of the vibration waveform by the absolute value and the LPF processing is performed next, and the frequency spectrum of the waveform subjected to the envelope processing is calculated. Next, the bearing path frequency component of f _inn. F _out. F _ball is extracted.
Based on these measurement results, the cause of the abnormality is estimated in consideration of the magnitude of each vibration component. Causes of this abnormality include imbalance and misalignment of rolling bearings and loosening of the foundation.

  The acoustic emission method is a method for diagnosing an early detection of a failure of a rolling bearing and a remaining life using an AE signal having a frequency higher than that of acceleration. This AE method is a diagnostic method using an AE signal, which is a phenomenon in which distortion energy stored up to that time propagates as sound when an object is deformed or destroyed. This AE signal is the propagation of an elastic wave when the elastic energy inside the material is released, and is not necessarily limited to when it breaks, but also covers dislocations and transformations of the crystal structure of the material. The rolling bearing rotating the AE signal is subjected to signal processing using an AE sensor, and the occurrence frequency of the AE wave is observed to diagnose the rolling bearing.

  Using such a diagnostic method, an unexpected failure of the rolling bearing is predicted in advance, and the replacement time of the bearing is predicted in advance. In this way, the “accident-free life” for recognizing bearing anomalies and the “accident life” that leads to seizure and breakage of the bearing are clarified, and the duration of the accident-occurring life, that is, the remaining life is predicted from the accident-free life. ing. As described above, conventionally, the presence / absence of an abnormality in a rotating device and the cause estimation are performed, the degree of the abnormality is determined, and the repair timing of the rolling bearing is determined. In the most frequently used statistical prediction of acceleration vibration, the vibration value up to the point of life prediction is used as a parameter, and the curve is regressed to a quadratic curve or exponential function, and the remaining life is obtained by obtaining the period until reaching an acceptable vibration value. .

  However, in the conventional diagnosis or prediction method described above, it is difficult to set an acceptable vibration value, and the remaining life varies greatly depending on the setting of the value, and it is difficult to predict the remaining life with high accuracy. Also, when the vibration starts to increase, it is already at the end of the bearing life, so it is difficult to make a long-term maintenance plan, and repair may not be in time even if the remaining life is predicted. For this reason, in reality, there are many cases where the rolling bearing is replaced at an early stage even though there is a sufficient margin for the true life. In addition, due to the low accuracy of remaining life diagnosis, in actual power plants, factories, etc., it is not possible to extend the bearing inspection cycle, and a maintenance system called periodic inspection in which all bearings are replaced every few years. Had to be forced. These have problems such as reduction of maintenance original cost and hindering labor saving.

  Further, the conventional acoustic emission method can be diagnosed earlier than the statistical method using acceleration, but the AE (Acoustic Emission) sensor and signal processing circuit used for the diagnosis are expensive. Moreover, since the AE wave is delicate, it has a problem that it is easy to pick up ambient noise.

  The present invention has been developed to solve such problems. That is, the object of the present invention is to use the resonance frequency band signal and high frequency signal of the acceleration sensor to indicate the deterioration state of the lubricating oil due to the mixing of dust and water into the lubricating oil, which greatly affects the life of the rolling bearing. Accordingly, it is an object of the present invention to provide a method for diagnosing the remaining life of a rolling bearing, which can be detected at low cost and the life of the rolling bearing can be estimated with high accuracy at an early stage based on the detected state of dust and lubricating oil.

The remaining life diagnosis method of the present invention relates to the relationship between dust contamination and vibration / bearing life in the rolling bearing (3), and the deterioration and vibration / bearing life of the lubricant, and the model number and manufacturer of each rolling bearing (3). For each bearing specification such as a name, a vibration signal is obtained using an acceleration sensor (4), and a basic data collection stage for collecting a frequency band signal including a resonance frequency band signal capable of detecting the highest sensitivity by an experimental device; Resonance that can be detected with the highest sensitivity by using the acceleration sensor (4) for the rolling bearing (3) for diagnosing the remaining life of the rotating equipment (1, 2) such as pumps and fans. Using a measurement stage for measuring a frequency band signal including a frequency band signal, a measurement value obtained in the measurement stage, a determination result for the bearing specifications, and data obtained in the basic data collection stage, Serial to estimate the deterioration state of the dust mixed state and the lubricating oil of the diagnostic rolling bearing (3), a determination step for calculating a remaining life of該被diagnostic rolling bearing (3), formed Ri from the determination step, the For the rolling bearing to be diagnosed (3), the vibration waveform of 1 k to 64 kHz is divided into frequencies by a 1/2 octave band, a plurality of bands are sampled, the envelope processing is performed on the waveform of each frequency band, and the frequency Calculate the spectrum, determine the bearing specifications such as the bearing model number and manufacturer name, and when the model number and manufacturer name of the diagnosis rolling bearing (3) are known, calculate the bearing path frequency from the bearing specifications, The path frequency components ( three of f _inn. F _out. F _ball ) are extracted from the envelope processing spectrum of each frequency band, and the relative sensitivity (ratio to normal) of the extracted path frequency components of each frequency band is calculated. Pass frequency component Obtain the average of the upper band of the number of bands divided by relative sensitivity, determine whether any of the upper band averages of the number of bands divided into multiple bands of the pass frequency component exceeded the threshold, and exceeded the threshold When it is not, it is estimated that it is normal .

  For example, in the determination step, as a result of processing the vibration measurement waveform of the diagnosis rolling bearing (3), the calculated vibration value and the calculated life of the rolling bearing (3) are used to calculate the remainder of the rolling bearing (3). Calculate lifespan.

  In the determination step, for the diagnostic rolling bearing (3), a vibration waveform of 5k to 35kHz is divided by a frequency of 1/2 octave band to obtain a total of 6 bands, and an envelope is obtained for each frequency band waveform. Line processing is performed and a frequency spectrum is calculated.

  The threshold value in the determination step is a numerical value exceeding 1.

The remaining life diagnosis method according to the detection method of the lubricant deterioration of the present invention is based on the relationship between the contamination state of the rolling bearing (3) and the vibration / bearing life, and the acceleration sensor ( 4) is used to obtain vibration signals, and a basic data collection stage for collecting frequency band signals including resonance frequency band signals that can be detected with the highest sensitivity using an experimental device, and rotating devices such as pumps and fans (1, 2) Measurement of a rolling bearing (3) provided for the purpose of diagnosing the remaining life by obtaining a vibration signal using an acceleration sensor (4) and measuring a frequency band signal including a resonance frequency band signal capable of detecting the highest sensitivity. Using the measured value obtained in the step, the measured value obtained in the measuring step, and the data obtained in the basic data collecting step, the contamination state of the diagnostic rolling bearing (3) and the deterioration state of the lubricating oil are estimated. A determining step for calculating the remaining service life of該被diagnostic rolling bearing (3), Ri consists, in the above determination step, the for the diagnostic rolling bearing (3), the number of bands of the upper divided into a plurality band-pass frequency components Determine whether any of the band averages exceeded the threshold value, and when the threshold value was exceeded, create a wavelet distribution of the vibration waveform to prevent misidentification of indentation / lubricant deterioration, and time The frequency distribution (wavelet) is displayed, and the time frequency distribution (wavelet) by humans is confirmed .

Or, use the acceleration sensor (4) to obtain vibration signals for the relationship between dust contamination in the rolling bearing (3) and vibration / bearing life, and the deterioration and vibration / bearing life of the lubricating oil. A basic data collection stage for collecting frequency band signals including resonant frequency band signals that can be used by an experimental device, and rolling bearings (3) provided in rotating devices (1, 2) such as pumps and fans to diagnose the remaining life ) For the vibration signal using the acceleration sensor (4), and measuring the frequency band signal including the resonance frequency band signal that can be detected with the highest sensitivity, the measurement value obtained by the measurement stage, Using the data obtained in the basic data collection stage, the state of dust contamination and the deterioration state of the lubricating oil of the diagnosed rolling bearing (3) is estimated, and the remaining life of the diagnosed rolling bearing (3) is calculated. A constant phase, formed Ri from, in the determination step, the for the diagnostic rolling bearing (3), the relative sensitivity average of the upper three bands of the peak frequency components determined whether the threshold has been exceeded, threshold In order to prevent misidentification of indentation / lubricating oil deterioration, the wavelet distribution of the vibration waveform is created, the time frequency distribution (wavelet) is displayed, and the human time frequency distribution (wavelet) is confirmed. It is characterized by that.

Furthermore, the acceleration signal (4) is used to obtain the vibration signal for the relationship between dust contamination in the rolling bearing (3) and vibration / bearing life, and the deterioration of the lubricant and vibration / bearing life. A basic data collection stage for collecting frequency band signals including resonant frequency band signals that can be used by an experimental device, and rolling bearings (3) provided in rotating devices (1, 2) such as pumps and fans to diagnose the remaining life ) For the vibration signal using the acceleration sensor (4), and measuring the frequency band signal including the resonance frequency band signal that can be detected with the highest sensitivity, the measurement value obtained by the measurement stage, The data obtained in the basic data collection stage is used to estimate the contamination state of the diagnosis rolling bearing (3) and the deterioration state of the lubricating oil, and to calculate the remaining life of the diagnosis rolling bearing (3). A step, formed Ri from, in the determination step, the for the diagnostic rolling bearing (3), broadband RMS value relative sensitivity × narrowband effective value relative sensitivity, determines whether the threshold has been exceeded, threshold In order to prevent misidentification of indentation / lubricating oil deterioration, the wavelet distribution of the vibration waveform is created, the time frequency distribution (wavelet) is displayed, and the human time frequency distribution (wavelet) is confirmed. It is characterized by that.

  In the determination step, the remaining life of the indentation start type peeling mode and the lubricant deterioration mode is diagnosed using a remaining life estimation formula calculated as a product of a calculated life and a remaining life coefficient of 0 to 1.

In the determination step, for the diagnostic rolling bearing (3), it is determined whether any of the upper band averages of the number of bands divided into a plurality of bands of the pass frequency component has exceeded a threshold, and the threshold has been exceeded. The indentation origin type peel mode is confirmed, the indentation size is estimated from the relative sensitivity average of the top three bands, and the remaining life in the indentation origin type peel mode is expressed by the following formula [Equation (1)]. 5. The remaining life diagnosis method for a rolling bearing according to claim 4, wherein the diagnosis is performed by calculation.

In the determination step, the remaining life estimation formula in the indentation-origin type peeling mode is calculated life as the basic dynamic rating life of the following formula [Equation (2)], and the remaining life coefficient is a function of P / C and vibration value. Remaining life from indentation occurrence.

In the determination step, the remaining life estimation formula in the lubricant deterioration mode is the calculated life, the calculated life is the calculated life by the following formula [Equation (3)], and the remaining life coefficient is a function of the vibration value. Remaining life from occurrence.

In the determination step, for the diagnostic rolling bearing (3), an effective value of the sensor resonance frequency band 23k to 32 kHz and a relative sensitivity of 5k to 35 kHz effective value are calculated, and 23k to 32 kHz effective value relative sensitivity × 5 k to 35 kHz effective. It is determined whether the value relative sensitivity exceeds the threshold value, and when the value exceeds the threshold value, it is confirmed that it is in the lubricant deterioration mode, and the effective value relative sensitivity of 23 k to 32 kHz is used in the lubricant deterioration mode. The remaining life is calculated by the following formula [Equation (4)] and diagnosed.

  In the determination step, a cepstrum having a vibration waveform of 1 K to 64 kHz is calculated for the diagnostic rolling bearing (3), a cepstrum coutosis is calculated, and it is determined whether the cepstrum coutosis exceeds a threshold, When the value exceeds the value, it is determined that the indentation start type peeling mode is selected, and when the value does not exceed the threshold value, the lubricant deterioration mode is estimated.

  The threshold value in the determination step is a numerical value exceeding 3.

  In the determination step, if it is determined that there is an abnormality in the indentation start type peeling mode or the lubricant deterioration mode, the determination by the human is unnecessary and the time frequency distribution (wavelet) is not displayed.

  In the determination step, when it is determined that the diagnostic rolling bearing (3) is in the indentation start type peeling mode when the threshold value is exceeded, the indentation size is estimated from the relative sensitivity average of the upper three bands. Then, the remaining life in the indentation origin type peeling mode is diagnosed, and the remaining life is expressed in units of time.

  In the determination step, a vibration waveform of 5 k to 35 kHz is measured for the diagnostic rolling bearing (3). In order to detect the indentation, the envelope processing spectrum is calculated by dividing the 5 k to 35 kHz band into six frequency bands, and it is determined whether or not the envelope processing spectrum exceeds the threshold value. If not, it is estimated to be normal, and if the threshold value is exceeded, it is estimated that there is an abnormality although the indentation origin type peeling mode abnormality / failure mode cannot be specified.

  In the determination step, in order to detect the lubricant deterioration of the diagnostic rolling bearing (3), the feature amount (23 k to 32 kHz execution value × 5 k to 35 kHz execution value) is calculated, and the feature amount of the lubricant deterioration is calculated. It is judged whether the threshold value has been exceeded. If the threshold value is not exceeded, it is assumed that the threshold value is normal.If the threshold value is exceeded, the lubricant deterioration mode failure / failure mode cannot be specified but is abnormal. Estimate that there is.

  The threshold value in the determination step is 2.0.

  In the determination step, the feature amount (cepstrum cootsis) is calculated in order to discriminate between the indentation and the lubricant deterioration for the roller bearing (3) to be diagnosed, and it is determined whether or not the cepstrum cotsis exceeds a threshold value. When the threshold value is exceeded, it is determined that there is an abnormality in the indentation start type peeling mode, and when the threshold value is not exceeded, it is estimated that there is an abnormality in the lubricant deterioration mode.

  The threshold value in the determination step is 3.8.

  In the determining step, when the indentation detection result of the diagnosed rolling bearing (3) does not exceed the threshold value and the result of the lubricant deterioration detection also does not exceed the threshold value, the diagnosed rolling bearing ( It is determined that 3) is normal.

  In the determination step, when the result of the indentation detection exceeds the threshold value for the diagnostic rolling bearing (3) and the failure mode cannot be specified, or the result of the lubricant deterioration detection exceeds the threshold value. When the failure mode cannot be specified, it is recognized that it is a caution. The indentation amount, the lubricant deterioration amount, the indentation / lubricant identification amount are specified, and the time frequency distribution (wavelet) is displayed. The failure mode is determined and diagnosed by a human.

  In the determination step, when the result of indentation detection exceeds the threshold value for the diagnostic rolling bearing (3), and the cepstrum cootosis exceeds the threshold value, the indentation origin type peeling mode is set. It is determined that there is an abnormality and indentation is generated, and the remaining life in the indentation origin type peeling mode is diagnosed.

  If it is determined in the determination step that the indentation start-type separation mode of the diagnostic rolling bearing (3) is abnormal, or if it is determined that there is an abnormality in the lubricant deterioration mode, it is recognized that attention is required, and the amount of indentation, lubrication The amount of oil deterioration and the indentation / lubricant identification amount are specified, the time frequency distribution (wavelet) is displayed, and the failure mode is determined and diagnosed by a human.

  In the determination step, threshold values for detecting indentation and lubricant deterioration are (a) BRG size, (b) equipment weight, (c) motor output, (d) equipment type, (e) these (a) It can change by the combination of (d).

  In the judgment stage, in addition to detection of indentation and lubricant deterioration, all faults of rotating machines are added, including conventional fault detection methods that diagnose equipment unbalance, misalignment, loose foundations, etc. by frequency analysis. Can be diagnosed automatically.

In the diagnosis method having the above-described configuration, the relationship between the acceleration and the dust contamination state of the indentation state of the rolling bearing generated when the lubrication due to the contamination of the dust or the lubricant is deteriorated in advance in the basic data collection stage. The basic data collected by the experimental device is acquired regarding the relationship between the acceleration and the lubricating oil state. In the measurement stage, the vibration signal is obtained using the acceleration sensor (4) for the diagnostic rolling bearing (3) provided in the rotating device (1, 2) to be diagnosed for the remaining life, and the most sensitive detection is performed. Measure possible resonant frequency band signals or high frequency signals.
Next, in the determination stage, the measured value obtained in the measurement stage, the data obtained in the basic data collection stage, the bearing load, the rotational speed, the operation time, and the rolling in the normal state of the diagnostic rolling bearing (3) measured in advance. Using the vibration data relating to the bearing identification number, the contamination state of the diagnosis rolling bearing (3) and the deterioration state of the lubricating oil are estimated, and the remaining life of the diagnosis rolling bearing (3) is calculated.

In this determination step, first, it is determined whether the diagnosis rolling bearing (3) is in the initial stage of deterioration or in the final stage by calculating the increasing tendency of vibration.
When it is determined that the deterioration is in the initial stage, it is further determined as follows. First, the resonance frequency band signal or the high frequency band signal of the acceleration sensor (4) of the diagnosis rolling bearing (3) obtained in the measurement step, and the vibration data in the normal state of the diagnosis rolling bearing (3) measured in advance. It is used to determine whether the diagnostic rolling bearing (3) is in a normal deterioration process, a dust mixing process, or a lubricant deterioration process.

If it is determined that dust is not mixed in the roller bearing (3) to be diagnosed, the lubricating oil is not in a deteriorated state, and the bearing state is normal at the initial stage of deterioration, the rated life is calculated as the remaining life. . Next, when it is determined that dust has entered the roller bearing (3) to be diagnosed and is in an initial deterioration state, the size of the mixed dust is estimated from the vibration data in the basic data collection stage, and the remaining life is estimated. calculate. Furthermore, when the lubricating oil of the diagnostic rolling bearing (3) has deteriorated, it is determined that the bearing state is in the initial stage of deterioration, the deterioration of the lubricating oil is estimated from vibration data at the data collection stage, and the remaining life is estimated. calculate.

  Finally, when it is determined that the deterioration is at the end, the remaining life is calculated based on the increasing tendency of the low-frequency vibration of the acceleration. At this time, in the diagnostic method of the present invention, more accurate remaining life is calculated by observing the elapsed time from dust contamination or deterioration of the lubricant to the rapid increase in acceleration as well as predicting the increase tendency of vibration. be able to.

  As described above, the present invention determines the replacement time by estimating the life of the rolling bearing (3) during the operation of the rotating device (1, 2), thereby more efficiently maintaining the device (1, 2). Is possible. For example, with regard to rotating equipment (1, 2) in a power plant, it is easy to make a plan to replace the rolling bearing (3) in the fall, avoiding summer when the operating rate is high, and improve maintenance efficiency. Can do. In addition, since it is possible to perform early residual life diagnosis for rotating equipment that had to be regularly inspected due to insufficient accuracy of the conventional remaining life diagnosis method, it becomes easier to extend the inspection cycle and collect deterioration data, It can be expected to improve the efficiency of the transition from the conventional periodic inspection system to the state-based maintenance in which maintenance is performed according to the equipment.

  In particular, in the present invention, the remaining life is accurately determined based on the characteristics of the bearing by determining the bearing specifications such as the model number and manufacturer name of the roller bearing (3) to be diagnosed for which the remaining life is to be diagnosed. Can be diagnosed. That is, based on whether or not the model number of the rolling bearing (3) to be diagnosed and the manufacturer name can be discriminated, it is determined whether or not the threshold value has been exceeded from the subsequent specific measurement result, and whether or not normal or abnormal is determined. Can be estimated.

In the present invention, the remaining life is accurately determined by determining whether the abnormality of the roller bearing (3) to be diagnosed is due to an indentation or due to an abnormality in running out of lubricating oil. For example, when it is estimated that the indentation origin type peeling mode is used, the remaining life can be diagnosed by estimating the indentation size based on a predetermined mathematical formula. Further, when it is estimated that the lubricant deterioration mode is set, the remaining life can be diagnosed by estimating the degree of lubricant deterioration based on a predetermined mathematical formula.
When it is difficult to determine whether there is an abnormality due to the indentation-origin peeling mode or the lubricant deterioration mode, a reliable remaining life can be ensured by checking the time frequency distribution (wavelet) by humans to prevent misidentification. Can be diagnosed.

  As described above, by comprehensively determining both the indentation start type peeling mode and the lubricant deterioration mode, a more precise remaining life can be diagnosed for the rolling bearing (3). The remaining life diagnosis method of the present invention can be combined with a conventional detection method for detecting conventional imbalance, misalignment, and foundation looseness.

  As described above, the present invention estimates the life of the rolling bearing (3) during operation of the rotating device (1, 2), so that the inspection schedule / diagnostic report output unit (10) has a replacement period or replacement time. This makes it possible to maintain the equipment more efficiently. For example, with respect to the rotating equipment (1, 2) in the power plant, it is possible to easily make a plan to replace the rolling bearing (3) in the fall while avoiding the summer when the operation rate is high. In addition, it is possible to quickly determine the remaining service life of equipment that had previously had to be regularly inspected, which makes it easier to extend the inspection cycle and collect deterioration data. It is possible to streamline the transition to a state-based maintenance system that performs maintenance according to the state.

It is a block diagram which shows the remaining life diagnostic method of the rolling bearing of this invention. It is sectional drawing which shows an example of the rolling bearing with which the electric motor used as the target object diagnosed with the remaining life diagnostic method and the rotary equipment were equipped. It is a flowchart which shows the remaining life diagnostic method of a rolling bearing. It is a flowchart which shows the remaining life diagnosis method of the rolling bearing which determines a bearing specification. It is a flowchart which shows the remaining life diagnostic method of a rolling bearing by an indentation feature-value when bearing specifications are known. It is a flowchart which shows the remaining life diagnostic method of a rolling bearing by an indentation feature-value when a bearing specification is not known. It is a graph which shows the influence which the magnitude | size of an indentation and the magnitude | size of a load have on the remaining life of a bearing. It is a graph which shows the experimental result which calculated | required the relationship between the relative vibration value of a bearing by the lubricating oil running out, and the remaining life. It is a graph which shows the relationship between relative sensitivity and indentation size. It is the result of diagnosing the remaining life of the rolling bearing used in the motor for driving the pump, and shows the wavelet distribution when the waveform, FFT spectrum and enlarged from the upper stage, (a) is normal state, (b) is The state of occurrence of indentation and (c) shows the state in which the deterioration of the lubricating oil has occurred. It is a block diagram which shows the remaining life diagnostic apparatus of the rolling bearing of this invention. It is a flowchart which shows the conventional remaining life diagnostic method.

  The method for diagnosing the remaining life of a rolling bearing according to the present invention includes a basic data collection stage for collecting in advance a relationship between dust mixed state / deterioration state of lubricating oil and vibration / life in an experimental machine, and a rotary device such as a pump and a fan or an electric motor. A measurement stage for measuring a resonance frequency band signal or a high-frequency signal of a rolling bearing for diagnosing the remaining life of the rolling bearing provided in the rotating mechanism part, etc., a determination stage for determining the remaining life of the rolling bearing to be diagnosed, It is the method which consists of.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a remaining life diagnosis method for a rolling bearing according to the present invention. FIG. 2 is a cross-sectional view showing an example of a rolling bearing provided in an electric motor and a rotating device that are objects to be diagnosed by the remaining life diagnosis method.

  At the basic data collection stage, in order to simulate the entry of dust, scratches are made directly on the rolling surface of the disassembled bearing to generate indentations on the bearing, and less lubricating oil is used to simulate the deterioration of the lubricating oil. The basic data is collected by performing a test with a bearing load tester using the selected bearing. In addition, other methods for simulating the mixing of dust include mixing a foreign substance in place of dust into the lubricating oil, changing the amount and size of the mixed foreign substance, and changing the hardness of the mixed foreign substance. Similarly, as a method for simulating the deterioration of the lubricating oil, there is a method of using a lubricating oil that has been oxidatively deteriorated or mixing water.

  In this basic data collection stage, data on the relationship between the acceleration and the size of the indentation is acquired in advance for the formation state of the indentation on the rolling bearing 3 when the lubrication due to the mixing of dust or the deterioration of the lubricating oil has deteriorated. The main deterioration modes of the rolling bearing 3 are two deterioration modes, that is, an internal starting type peeling and a surface starting type peeling. In this internal origin type separation, the repetitive stress received by the rolling element rolling surface is concentrated below the surface of the rolling surface, and peeling occurs from the inside of the rolling surface. In the surface-origin type peeling, the surface of the rolling surface is scratched by the entry of foreign matter such as dust into the lubricating oil, and peeling occurs from the surface of the rolling surface. The original life of the bearing is the life in the internally-originated delamination mode, and this life has been extended several times to several tens of times the rated life of the bearing due to recent advances in material technology. On the other hand, the life of the surface-origin type delamination mode due to foreign matters mixed in the lubricating oil is remarkably shortened to a fraction to a tenth of the internal origin-type delamination life.

  As described above, the rolling bearing 3 has various deterioration modes, and it is very important to consider these deterioration modes and failure mechanisms in the remaining life diagnosis of the bearing. Therefore, in the present invention, in consideration of such a deterioration mode of the rolling bearing 3, the basic data collection stage is used as a precondition for diagnosing the remaining life that can be diagnosed earlier and with higher accuracy than in the past. .

  For the diagnostic rolling bearing 3 provided in the rotating devices 1 and 2 to be diagnosed for the remaining life, the remaining life diagnosis preparation stage and the measurement stage are taken out of the determination stages. In the remaining life diagnosis preparatory stage, the data on the bearing load, the rotational speed, the operation time, the rolling bearing identification number, and the vibration data at the normal time are collected for the diagnosis rolling bearing 3 using the acceleration sensor 4. In the measurement stage, a vibration signal is obtained for the diagnostic rolling bearing 3 during operation by using the acceleration sensor 4, and a resonance frequency band signal or a high frequency band signal that can be detected with the highest sensitivity is measured.

FIG. 3 is a flowchart showing a remaining life diagnosis method for a rolling bearing.
Using the data obtained in the basic data collection stage and the remaining life diagnosis preparation stage, it is estimated based on the flow as shown in FIG. 3 whether the roller bearing 3 to be diagnosed is in the initial stage of deterioration or in the final stage. Diagnose the remaining life.
A vibration waveform of 1 k to 64 kHz is measured for the diagnostic rolling bearing 3. Preferably, a vibration waveform of 5 k to 35 kHz is measured. In order to detect an indentation, an envelope processing spectrum is calculated by dividing the frequency band into a plurality of frequency bands, and whether the envelope processing spectrum of the plurality of frequency bands exceeds a threshold value (= 2.0). When the threshold value is not exceeded, it is estimated that it is normal. On the other hand, when the threshold value is exceeded, it is determined that there is an abnormality in the indentation start type peeling mode, or when it is abnormal but the failure mode cannot be specified, the process proceeds to the next determination.
For example, as a result of processing the vibration measurement waveform of the diagnosis rolling bearing 3, the remaining vibration life of the rolling bearing 3 is calculated and determined using the calculated vibration value and the calculated life of the rolling bearing 3.

  In order to detect lubricant deterioration, the feature amount (23k to 32 kHz execution value × 5 k to 35 kHz execution value) is calculated, and whether or not the feature amount of lubricant deterioration exceeds a threshold value (= 2.0). Judgment is made, and when the threshold value is not exceeded, it is estimated that it is normal. On the other hand, when the threshold value is exceeded, it is estimated that there is an abnormality in the lubricant deterioration mode, or the process proceeds to the next determination when it is abnormal but the failure mode cannot be specified.

  Here, the threshold value is preferably 2.0, but any value exceeding 1.0 can be used for determination. This threshold value is varied based on the size and type of the rolling bearing 3, the weight of the equipment used, the kind of equipment, the capacity (kw) of the electric equipment, and the like. Alternatively, it is varied based on these combinations.

  In order to distinguish between indentation and lubricant deterioration, the feature value (cepstrum cootsis) is calculated, it is judged whether the cepstrum cootsis exceeds a threshold value (= 3.8), and the threshold value is exceeded. It is determined that there is an abnormality in the indentation origin type peeling mode. On the other hand, when the threshold value is not exceeded, it is estimated that there is an abnormality in the lubricant deterioration mode.

  Here, the threshold value is preferably 3.8, but any value exceeding 3.0 can be used for determination. As described above, this threshold value is varied based on the size and type of the rolling bearing 3, the weight of the device used, the type of device, the capacity (kw) of the electric device, and the like. Alternatively, it is varied based on these combinations.

  When the result of the indentation detection does not exceed the threshold value, and when the result of the lubricant deterioration detection also does not exceed the threshold value, it is determined that the diagnostic rolling bearing 3 is “normal”. Even when the result of indentation detection does not exceed the threshold value, when it is determined that the result of lubricant deterioration detection is abnormal, it is determined that the roller bearing 3 to be diagnosed is abnormal. Similarly, even when the result of detection of lubricant deterioration does not exceed the threshold value, when it is determined that the result of indentation detection is abnormal, it is determined that the roller bearing 3 to be diagnosed is abnormal.

  When the indentation detection result exceeds the threshold and the failure mode cannot be specified, or when the lubricant deterioration detection result exceeds the threshold and the failure mode cannot be specified When it is necessary to be careful, the indentation amount, the lubricant deterioration amount, the indentation / lubricant identification amount are specified, and the time frequency distribution (wavelet) is displayed. Further, the failure mode is determined and diagnosed by a human.

  When the result of the indentation detection exceeds the threshold value and the cepstrum cootsis exceeds the threshold value, it is determined that there is an abnormality in the indentation origin type peeling mode. As a result, it is determined that the indentation is generated, and the remaining life in the indentation origin type peeling mode is diagnosed.

  When the result of the lubricant deterioration detection exceeds the threshold value and the cepstrum cootsis does not exceed the threshold value, it is determined that there is an abnormality in the lubricant deterioration mode. As a result, it is determined that the lubricant has deteriorated, and the remaining life in the lubricant deterioration mode is diagnosed.

  Furthermore, when it is estimated that there is an abnormality in the indentation origin type peeling mode, or when it is determined that there is an abnormality in the lubricating oil deterioration mode, it is recognized that it is necessary to pay attention, and the indentation amount, the lubricating oil deterioration amount, the indentation / lubricating oil The identification amount is specified, and the time frequency distribution (wavelet) is displayed. Further, the failure mode is determined and diagnosed by a human.

  In addition, when it is determined that there is an abnormality in the indentation start type peeling mode or the lubricant deterioration mode, the indentation amount, the lubricant deterioration amount, the indentation / lubricant identification amount is not specified, and the time frequency distribution (wavelet) Do not show.

FIG. 4 is a flowchart showing a method for diagnosing the remaining life of a rolling bearing for determining the model number and manufacturer name of the rolling bearing according to the second embodiment. FIG. 5 is a flowchart showing a method for diagnosing the remaining life of a rolling bearing based on the indentation feature amount when the bearing specifications are known. FIG. 6 is a flowchart showing a method for diagnosing the remaining life of a rolling bearing based on indentation features when the bearing specifications are unknown.
A description will be given of an indentation detection method that involves determination of specifications such as the model number and manufacturer name of the rolling bearing 3 of the second embodiment.
For the rolling bearing 3 to be diagnosed, the vibration waveform of 5k to 35kHz is divided by 1/2 octave band and a total of 6 bands are sampled. Envelope processing is performed on the waveform of each frequency band, and the frequency spectrum is obtained. Calculate and determine the model number and manufacturer name of the rolling bearing 3. This is because if the model number and manufacturer name of the rolling bearing 3 can be discriminated, the characteristics of the bearing can be understood and the remaining life can be estimated more accurately.

When the model number and manufacturer name of the rolling bearing are known, the following determination is made.
The bearing path frequency is calculated from the specifications of the rolling bearing 3, the path frequency components (three of f _inn. F _out. F _ball ) are extracted from the envelope processing spectrum of each frequency band, and the path of each extracted frequency band The relative sensitivity of the frequency component (ratio to normal) is calculated, and the upper three band average of the relative sensitivity of the pass frequency component is obtained. The calculation is performed independently for three of f _inn. f _out. f _ball . It is determined whether any of the upper three band averages (three of f _inn. F _out. F _ball ) of this band frequency component has exceeded the threshold value (= 2.0), and the threshold value has not been exceeded. When it is estimated to be normal.

  On the other hand, when this threshold value is exceeded, human time frequency distribution (wavelet) is checked to prevent misidentification of indentation / lubricating oil deterioration, wavelet distribution of vibration waveform is created, and time frequency distribution ( Wavelet) is displayed. This is because human judgment can be most accurately performed.

If you do not know the rolling bearing model and manufacturer name, make the following judgment.
For the diagnostic rolling bearing 3, the peak of the frequency spectrum is extracted within the range of the rotation speed × 1.5 to 7 from the envelope processing spectrum of each frequency band, and the relative sensitivity of the peak frequency component of each extracted frequency band ( (Compared with normal)), the average of the upper three bands of the relative sensitivity of the peak frequency component is calculated, and whether the average of the upper three bands of the peak frequency component exceeds the threshold (= 2.0) When the threshold value is not exceeded, it is estimated that it is normal.

  On the other hand, when the threshold value is exceeded, it is confirmed that it is an indentation origin type peeling mode, the indentation size is estimated from the relative sensitivity average of the top three bands, and the remaining life in the indentation origin type peeling mode is diagnosed. The remaining life is calculated by the following formula [Equation 1] and diagnosed.

As described above, according to the present invention, the life of the rolling bearing 3 is estimated during the operation of the rotating devices 1 and 2, so that the replacement time is determined, and more efficient maintenance of the rotating devices 1 and 2 is possible. For example, with respect to the rotating devices 1 and 2 in the power plant, it is possible to easily make a plan to replace the rolling bearing 3 in the autumn while avoiding the summer when the operation rate is high.
A test bearing that uses a Rockwell hardness tester on the inner ring of a rolling bearing to create a conical indentation with an indenter, conduct a fatigue test with varying indentation size and load load, and determine the operating time after the indentation occurs. Is a JIS6206 (inner diameter 30 mm, outer diameter 62 mm, width 16 mm) tested using a fatigue testing machine manufactured by the Japan Society for the Promotion of Science 126 using deep groove ball bearings that are most commonly used in motors. . The rotational speed is 2000 rpm, and the lubricating oil is turbine oil ISOVGIO. As shown in FIG. 7 , it was found that the life ratio of the rolling bearing to the calculated life decreases as the size of the indentation increases, and the decrease rate of the life ratio increases as the load load decreases.

FIG. 7 is a graph showing the effect of indentation size and load size on the life.
Here, P is the radial load at the time of the bearing test, C is the basic dynamic load rating of the bearing, and LlOh is the basic dynamic rating life of JISB1518 / 1992 as shown in Formula [Equation 1]. L10 is a 10% life when the life test data is applied to the Weibull distribution.

Here, n is the rotational speed (rpm), C is the basic dynamic load rating of the bearing, and P is the radial load during the bearing test. The basic dynamic rating life of a rolling bearing is the time that 90% of the same group of bearings can be rotated one million times without causing flaking due to rolling fatigue when they are individually rotated under the same conditions. is there.
From FIG. 8, the mathematical formula [Equation 2] can be confirmed in each indentation of each size.

According to the study of Sayles et al. (Reference: Sayles et al., ASTM STP771 (1982) 255-274), a roller of 3 microns or less is used to supply the rolling bearing. It has been found that the life is constant when the cleanliness of the lubricating oil is controlled. Further, the raceway surface of the rolling bearing is superfinished, and the surface roughness pitch is about 2.5 μm. Therefore, when the indentation size is 2.5 μm, the life of the rolling bearing when there is no indentation considered that there is no indentation due to the indentation is longer than the basic dynamic rating life due to the recent improvement in cleanliness of bearing steel. . The unified examination conducted by the Japan Society for the Promotion of Science is said to be twice as capable. Therefore, the relative life when the indentation size is 2.5 μm is set to 2.
The regression coefficients a and b were determined based on the experimental results when the indentation size was 230 μm and 460 μm and the results when the indentation size was 2.5 μm. As a result, the following formula 2 was obtained.

Here, d is the diameter of the indentation, and the unit is micron.
Equation 4 was obtained by modifying the equation to estimate the remaining life.

  The size of the indentation can be determined by processing the vibration waveform of the bearing.

When the lubricating oil film of the rolling bearing is broken, the oxidation deterioration of the lubricating oil is accelerated due to heat generation in the metal contact portion, and the oxidation deterioration of the lubricating oil is accelerated due to the catalytic action caused by the abrasion powder in the metal contact portion.
The lubrication state between the bearing ring of the rolling bearing and the rolling element and the bearing vibration have the relationship as described above, and the lubrication state was changed. A lubrication life test was conducted, and the relationship between the relative vibration value and the relative life was obtained. The test bearings and the test machine were indented. Same as in the test. The test load was 10% of the basic dynamic load rating. Tests were performed in dry state and with ISOVG2 lubricating oil. The rotation speed is 900 rpm and 1800 rpm. The results are shown in the figure, and an empirical formula was obtained between the relative vibration value with respect to the normal state and the relative life (ratio of the life of the test bearing to the calculated life obtained from Booster's grease life formula).

Next, a method for detecting deterioration of lubricating oil in a rolling bearing will be described.
For the diagnostic rolling bearing 3, the effective value of the sensor resonance frequency band 23k to 32kHz is calculated, the relative sensitivity of the 23k-32kHz effective value is calculated, and the relative sensitivity of the 23k-32kHz effective value is a threshold value (= 1.5). ) Is exceeded, and if the threshold value is not exceeded, it is estimated that it is normal.

On the other hand, when the threshold value is exceeded, it is confirmed that the indentation origin type peel mode is in effect, and the indentation size is estimated from the relative sensitivity average of the upper three bands. Diagnose the remaining life in the indentation-peeling mode. The remaining life is calculated by the following formula [Numerical formula (1)] and diagnosed.

  Alternatively, when the threshold value is exceeded, it is confirmed that the mode is the lubricant deterioration mode, the degree of lubricant deterioration is estimated from the effective value of 23 to 32 kHz, the remaining life is diagnosed in the lubricant deterioration mode, and the remaining life is determined. Display time unit (hrs) and diagnose.

Calculate cepstrum of 5k to 35kHz vibration waveform for the rolling bearing 3 to be diagnosed, calculate the cepstrum coutsis, determine whether the cepstrum coutsis exceeds a threshold value (= 3.8), and exceed the threshold value If it is, it is determined that the indentation origin type peel mode is selected.
The indentation size is estimated from the average of the relative sensitivity of the upper three bands, the remaining life in the indentation origin type peeling mode is diagnosed, and the remaining life is calculated by the following formula [Numerical formula (1)].

On the other hand, when the threshold value is not exceeded, it is determined that the mode is the lubricant deterioration mode, the degree of lubricant deterioration is estimated from the effective value of 23 to 32 kHz, the remaining life is diagnosed in the lubricant deterioration mode, and the remaining life is Diagnose by calculating with the formula [number (2)].

  FIG. 8 is a graph showing the relationship between the relative vibration value of the bearing due to running out of oil and the remaining life. The calculated lifetime in FIG. 8 is Lhb of the Booser equation of Formula [5].

Here, t is a bearing outer ring temperature (° C.), the half-life subtraction coefficient according to S _G grease composition, S _N and S _W is the half life subtraction coefficient according to each rotational speed, with a half-life subtraction coefficient according to the load and speed Each formula is as shown in [Equation 6] and [Equation 7].

  Here, d (mm) is the inner diameter of the bearing, nL is the catalog allowable rotational speed, and the unit is rpm.

Here, P (lbf) is a load load and Cr (lbf) is a basic dynamic load rating.
Therefore, the remaining life is expressed by the following equation (8).

  As described above, when the failure mode cannot be identified with respect to the diagnosis result when the model number and manufacturer name of the rolling bearing are determined, a wavelet distribution of the vibration waveform is created. Next, it recognizes that it is a caution, specifies the indentation amount, the lubricant deterioration amount, the indentation / lubricant identification amount, and displays the time frequency distribution (wavelet).

  When there is an abnormality in the indentation mode, the indentation size is estimated from the relative sensitivity average of the upper three bands. Diagnose the remaining life in the indentation origin type peeling mode and display the remaining life in hours.

  When there is an abnormality in the lubricating oil deterioration mode, the degree of deterioration of the lubricating oil is estimated from an execution value of 23 k to 32 kHz. Diagnose the remaining life in the lubricant deterioration mode and display the remaining life in hours.

FIG. 9 is a graph showing the relationship between relative sensitivity and indentation size.
In the present invention, the detection of the occurrence of indentation due to dust contamination and the size of the indentation in the basic data collection stage, the measurement stage, and the judgment stage, and the frequency including the vibration signal in the resonance frequency band where the acceleration sensor 4 can detect the highest sensitivity It can be determined by bandwidth. As shown in FIG. 10, the indentation size estimation curve can be changed according to the type of device and the size of hearing, and the indentation size can be estimated from the calculated relative sensitivity.

FIG. 10 shows the result of diagnosing the remaining life of the rolling bearing used in the motor for driving the pump. The waveform, FFT spectrum, and enlarged wavelet distribution are shown from the upper stage, and (a) shows the normal state, ( b) shows an indentation state, and (c) shows a state in which the lubricant has deteriorated.
If the life of a normal rolling bearing is assumed to be relative life 1 (the rated life is assumed to be relative life 1), the life of the bearing with the indentation is much shorter than that and may be 1/100 or less of the rated life. . Further, it is known that the peeling of the diagnostic rolling bearing 3 starts from an indentation without exception, and the variation in the bearing life is extremely small. The determination step of the present invention is estimated based on the relationship between the indentation size and the lifetime.

FIG. 11 is a block diagram showing a remaining life diagnosis device for a rolling bearing according to the present invention.
The remaining life diagnosis device for a rolling bearing includes an acceleration sensor 4, an analog / digital converter 5, a feature amount extraction unit 6, a measurement result database 7, a remaining life diagnosis unit 8, a diagnosis result display unit 9, An inspection schedule / diagnostic report output unit 10 and a transmission modem 11 are provided.

  The analog / digital converter 5 converts the data obtained by the acceleration sensor 4 with respect to the diagnosis rolling bearing 3 or the like to be diagnosed for the remaining life described above. The feature quantity extraction unit 6 extracts a vibration signal in a resonance frequency band in which the most sensitive detection is possible among vibration signals converted by the analog / digital converter 5.

  As described above, the measurement result database 7 includes the basic data obtained by collecting the relationship between the dust contamination state / deterioration state of the lubricating oil and the vibration / life in advance in the experimental machine, and the rotary device 1 such as the pump and fan, or the electric motor 2. Data on the bearing load, rotational speed, operating time and rolling bearing identification number of the rolling bearing to be diagnosed 3 to be diagnosed for the remaining life provided in the rotating mechanism portion, and vibration data of the diagnostic rolling bearing 3 in the normal state Collect and save.

  The remaining life diagnosis unit 8 uses the data mounted in the measurement result database 7, and based on the vibration signal of the diagnosis rolling bearing 3 extracted by the feature amount extraction unit 6, The mixture and the deterioration state of the lubricating oil are judged, and the remaining life is diagnosed.

  The diagnosis result display unit 9 displays the result of the remaining life diagnosis unit 9. The inspection schedule / diagnosis report output unit 10 outputs the next inspection schedule of the roller bearing 3 to be diagnosed and a report of the diagnosis result to the printers 1, 2, etc. based on the diagnosis result of the remaining life diagnosis unit 9. Is. As described above, the present invention estimates the life of the rolling bearing 3 during the operation of the rotating devices 1 and 2, thereby determining the replacement period or the replacement time in the inspection schedule / diagnostic report output unit 10, and more efficiently. Equipment can be maintained. For example, with respect to the rotating devices 1 and 2 in the power plant, it is possible to easily make a plan to replace the rolling bearing 3 in the autumn while avoiding the summer when the operation rate is high.

  In the remaining life diagnosis apparatus having the above configuration, the contamination state of dust or the deterioration state of the lubricant is detected at low cost by using the resonance frequency band signal or the high frequency signal of the acceleration sensor 4, and the state of the detected dust The life of the rolling bearing can be estimated with high accuracy based on the state of the lubricating oil.

The transmission modem 11 connects the waveform data and the diagnosis result to the Internet line. Thus, by connecting to the Internet line, the remaining life of the predetermined rolling bearing 3 can be easily estimated in a remote place.

  The present invention is not limited to the embodiment of the invention described above, and any configuration can be used as long as the diagnosis rolling bearing 3 estimates the remaining life and the remaining life of the diagnosis rolling bearing 3 is calculated. Of course, the present invention is not limited, and various changes can be made without departing from the scope of the present invention.

  In addition, data is collected on the relationship between the state in which dust is mixed or the state in which the lubricating oil has deteriorated and the acceleration and the size of the indentation, etc. By comparing / determining the life relationship data and the vibration signal of the diagnosed rolling bearing 3 extracted by the feature amount extraction unit 6, the state of contamination of the diagnosed rolling bearing 3 and the deterioration state of the lubricating oil are estimated, Of course, the structure for diagnosing the remaining lifetime is not limited to the illustrated configuration, and various modifications can be made without departing from the scope of the present invention.

  The method for diagnosing the remaining life of a rolling bearing according to the present invention is a rolling bearing used in pumps, fan rolling bearings, or motors that drive these devices, which are auxiliary machines used in chemical plants, steelworks and power plants. It can be used when performing the remaining life diagnosis.

1 Pump, fan 2 Electric motor 3 Rolling bearing (diagnostic rolling bearing)
4 Acceleration Sensor 5 Analog / Digital Converter 6 Feature Extraction Unit 7 Measurement Result Database 8 Remaining Life Diagnosis Unit 9 Diagnosis Result Display Unit 10 Inspection Schedule / Diagnosis Report Output Unit 11 Transmission Modem 12 Printer

Claims (27)

Regarding the bearing specifications such as the model number and manufacturer name of each rolling bearing (3) regarding the relationship between dust contamination and vibration / bearing life in the rolling bearing (3), and the deterioration and vibration / bearing life of the lubricating oil, Obtaining a vibration signal using the acceleration sensor (4), and collecting a basic data collection stage by using an experimental device to collect a frequency band signal including a resonance frequency band signal that can be detected with the highest sensitivity;
Resonance that can be detected with the highest sensitivity by using the acceleration sensor (4) for the rolling bearing (3) for diagnosing the remaining life of the rotating equipment (1, 2) such as pumps and fans. A measurement stage for measuring frequency band signals including frequency band signals;
Using the measurement value obtained in the measurement step, the determination result for the bearing specifications, and the data obtained in the basic data collection step, the contamination state of the roller bearing (3) to be diagnosed and the lubricating oil estimating the deterioration state, a decision step and calculates the remaining service life of該被diagnostic rolling bearing (3), formed Ri from
In the determination step, the vibration rolling bearing (3) to be diagnosed is obtained by dividing a vibration waveform of 1 k to 64 kHz by dividing the frequency by a 1/2 octave band, and collecting a plurality of bands. Perform line processing, calculate frequency spectrum, determine bearing specifications such as bearing model number and manufacturer name,
When the model number and manufacturer name of the diagnosis rolling bearing (3) are known,
The bearing path frequency is calculated from the bearing specifications, the path frequency components ( three of f _inn. F _out. F _ball ) are extracted from the envelope processing spectrum of each frequency band, and the extracted path frequency components of each frequency band Relative sensitivity (ratio to normal), and find the average of the upper bands of the number of bands divided by the relative sensitivity of the pass frequency component,
Determine whether any of the upper band averages of the number of bands divided into multiple bands of pass frequency components exceeded the threshold,
A method for diagnosing the remaining life of a rolling bearing, wherein the remaining life is estimated to be normal when the threshold value is not exceeded .   In the determination step, as a result of processing the vibration measurement waveform of the diagnostic rolling bearing (3), the remaining life of the rolling bearing (3) is calculated using the calculated vibration value and the calculated life of the rolling bearing (3). The remaining life diagnosis method for a rolling bearing according to claim 1, wherein the remaining life is calculated. In the determination step, for the diagnostic rolling bearing (3), a vibration waveform of 5k to 35kHz is divided by a frequency of 1/2 octave band to obtain a total of 6 bands, and an envelope is obtained for each frequency band waveform. performs line processing, calculates the frequency spectrum, remaining life diagnosis method of a rolling bearing according to claim 1, characterized in that. 2. The remaining life diagnosis method for a rolling bearing according to claim 1 , wherein the threshold value in the determination step is a numerical value exceeding 1. The most sensitive detection is possible by using the acceleration sensor (4) to determine the vibration signal regarding the relationship between dust contamination and vibration / bearing life in the rolling bearing (3) and the deterioration of the lubricant and vibration / bearing life. A basic data collection stage for collecting a frequency band signal including a resonance frequency band signal by an experimental device;
Resonance that can be detected with the highest sensitivity by using the acceleration sensor (4) for the rolling bearing (3) for diagnosing the remaining life of the rotating equipment (1, 2) such as pumps and fans. A measurement stage for measuring frequency band signals including frequency band signals;
Using the measured value obtained in the measurement step and the data obtained in the basic data collection step, the contamination state of the diagnostic rolling bearing (3) and the deterioration state of the lubricating oil are estimated, and the diagnostic rolling is performed. a determining step for calculating the remaining service life of the bearings (3), Ri consists,
In the determination step, for the diagnostic rolling bearing (3), it is determined whether any of the upper band averages of the number of bands divided into a plurality of bands of the pass frequency component has exceeded a threshold, and the threshold has been exceeded. When
Rolling bearing characterized by creating wavelet distribution of vibration waveform, displaying time frequency distribution (wavelet), and checking human time frequency distribution (wavelet) to prevent misidentification of indentation / lubricant deterioration Remaining life diagnosis method. The most sensitive detection is possible by using the acceleration sensor (4) to determine the vibration signal regarding the relationship between dust contamination and vibration / bearing life in the rolling bearing (3) and the deterioration of the lubricant and vibration / bearing life. A basic data collection stage for collecting a frequency band signal including a resonance frequency band signal by an experimental device;
Resonance that can be detected with the highest sensitivity by using the acceleration sensor (4) for the rolling bearing (3) for diagnosing the remaining life of the rotating equipment (1, 2) such as pumps and fans. A measurement stage for measuring frequency band signals including frequency band signals;
Using the measured value obtained in the measurement step and the data obtained in the basic data collection step, the contamination state of the diagnostic rolling bearing (3) and the deterioration state of the lubricating oil are estimated, and the diagnostic rolling is performed. a determining step for calculating the remaining service life of the bearings (3), Ri consists,
In the determination step, for the diagnostic rolling bearing (3), it is determined whether or not the relative sensitivity average of the upper three bands of the peak frequency component exceeds a threshold value.
To prevent misidentification of indentation / lubricant deterioration, create a wavelet distribution of vibration waveform, display the time frequency distribution (wavelet), and check the human time frequency distribution (wavelet). Bearing life diagnosis method. The most sensitive detection is possible by using the acceleration sensor (4) to determine the vibration signal regarding the relationship between dust contamination and vibration / bearing life in the rolling bearing (3) and the deterioration of the lubricant and vibration / bearing life. A basic data collection stage for collecting a frequency band signal including a resonance frequency band signal by an experimental device;
Resonance that can be detected with the highest sensitivity by using the acceleration sensor (4) for the rolling bearing (3) for diagnosing the remaining life of the rotating equipment (1, 2) such as pumps and fans. A measurement stage for measuring frequency band signals including frequency band signals;
Using the measured value obtained in the measurement step and the data obtained in the basic data collection step, the contamination state of the diagnostic rolling bearing (3) and the deterioration state of the lubricating oil are estimated, and the diagnostic rolling is performed. a determining step for calculating the remaining service life of the bearings (3), Ri consists,
In the determination step, for the diagnostic rolling bearing (3), it is determined whether or not the broadband effective value relative sensitivity × the narrow band effective value relative sensitivity exceeds a threshold value.
Rolling bearing characterized by creating wavelet distribution of vibration waveform, displaying time frequency distribution (wavelet), and checking human time frequency distribution (wavelet) to prevent misidentification of indentation / lubricant deterioration Remaining life diagnosis method.   In the determination step, diagnosing the remaining life of the indentation origin peeling mode and the lubricant deterioration mode using a remaining life estimation formula calculated as a product of a calculated life and a remaining life coefficient of 0 to 1. The method for diagnosing the remaining life of a rolling bearing according to claim 2. In the determination step, for the diagnostic rolling bearing (3), it is determined whether any of the upper band averages of the number of bands divided into a plurality of bands of the pass frequency component has exceeded a threshold, and the threshold has been exceeded. When
Confirm that it is indentation origin type peeling mode,
Estimate the indentation size from the average relative sensitivity of the top three bands,
The remaining life diagnosis method for a rolling bearing according to claim 1 , wherein the remaining life in the indentation start type peeling mode is calculated by the following formula [Numerical formula (1)].
In the determination step, the remaining life estimation formula in the indentation-origin type peeling mode is calculated life as the basic dynamic rating life of the following formula [Numerical formula (2)], and the remaining life coefficient is a function of P / C and vibration value. 9. The method for diagnosing the remaining life of a rolling bearing according to claim 8 , wherein the remaining life from the occurrence of indentation is used.
In the determination step, the remaining life estimation formula in the lubricant deterioration mode,
The calculated life, the calculated life the calculated life of Formulas [number (3)] below, claim 8, remaining life from the lubricant deterioration occurs as a function of the vibration value of the remaining service life coefficient be characterized by For remaining life of rolling bearings.
In the determination step, for the diagnostic rolling bearing (3), an effective value of the sensor resonance frequency band 23k to 32 kHz and a relative sensitivity of 5k to 35 kHz effective value are calculated, and 23k to 32 kHz effective value relative sensitivity × 5 k to 35 kHz effective. The value relative sensitivity determines whether the threshold is exceeded, and when it exceeds the threshold,
Confirm that it is in lubricant deterioration mode,
The rolling bearing according to claim 5, 6 or 7 , characterized in that the remaining life in the lubricant deterioration mode is calculated from the effective relative sensitivity of 23k to 32kHz by the following formula [Equation (4)]. Remaining life diagnosis method.
In the determination step, a cepstrum having a vibration waveform of 1K to 64 kHz is calculated for the diagnostic rolling bearing (3), a cepstrum coutsis is calculated, and it is determined whether the cepstrum coutsis exceeds a threshold value;
When the threshold value is exceeded, it is determined that the indentation origin type peeling mode is in effect,
Estimates that when the threshold value is not exceeded is the lubricant degradation mode, according to claim 1, 5, 6 or 7 remaining life diagnosis method of a rolling bearing of which is characterized in that. The method for diagnosing the remaining life of a rolling bearing according to claim 13 , wherein the threshold value in the determination step is a numerical value exceeding 3. In the determination step, when it is determined that there is an abnormality in the indentation origin peeling mode or the lubricant deterioration mode,
As human judgment is required, not to display the time-frequency distribution (wavelet), according to claim 1, 5, 6 or 7 remaining life diagnosis method of a rolling bearing of which is characterized in that. In the determination step, when it is determined that the diagnostic rolling bearing (3) is in the indentation start type peeling mode when the threshold value is exceeded,
4. The rolling bearing surplus according to claim 3 , wherein the indentation size is estimated from the relative sensitivity average of the top three bands, the remaining life in the indentation-origin type peeling mode is diagnosed, and the remaining life is expressed in units of time. Lifetime diagnosis method. In the determination step, a vibration waveform of 5 k to 35 kHz is measured for the diagnostic rolling bearing (3). In order to detect the indentation, the envelope processing spectrum is calculated by dividing the 5 k to 35 kHz band into six frequency bands, and it is determined whether or not the envelope processing spectrum exceeds the threshold value. If not, it is assumed to be normal,
The rolling bearing according to claim 1, 2, 5, 6, or 7, wherein when the threshold value is exceeded, an abnormality or failure mode of the indentation start type peeling mode cannot be specified but is abnormal. Remaining life diagnosis method. In the determination step, in order to detect the lubricant deterioration of the diagnostic rolling bearing (3), the feature amount (23 k to 32 kHz execution value × 5 k to 35 kHz execution value) is calculated, and the feature amount of the lubricant deterioration is calculated. Determine if the threshold has been exceeded, and if it does not exceed the threshold, assume that it is normal,
8. The method for diagnosing the remaining life of a rolling bearing according to claim 5, wherein when the threshold value is exceeded, the abnormality / failure mode of the lubricating oil deterioration mode cannot be specified but is estimated to be abnormal. .   The method for diagnosing the remaining life of a rolling bearing according to claim 17 or 18, wherein the threshold value in the determination step is 2.0. In the determination step, the feature amount (cepstrum cootsis) is calculated in order to discriminate between the indentation and the lubricant deterioration for the roller bearing (3) to be diagnosed, and it is determined whether or not the cepstrum cotsis exceeds a threshold value. And
When the threshold value is exceeded, it is determined that there is an abnormality in the indentation origin type peeling mode,
8. The remaining life diagnosis method for a rolling bearing according to claim 1 , wherein when the threshold value is not exceeded, it is estimated that there is an abnormality in the lubricant deterioration mode. 21. The remaining life diagnosis method for a rolling bearing according to claim 20 , wherein the threshold value in the determination step is 3.8. In the determining step, when the indentation detection result of the diagnosed rolling bearing (3) does not exceed the threshold value and the result of the lubricant deterioration detection also does not exceed the threshold value, the diagnosed rolling bearing ( The remaining life diagnosis method for a rolling bearing according to claim 1 , wherein 3) is determined to be normal. In the determination step, when the result of the indentation detection exceeds the threshold value for the diagnostic rolling bearing (3) and the failure mode cannot be specified, or the result of the lubricant deterioration detection exceeds the threshold value. When the failure mode cannot be specified, it is recognized that it is a caution. The indentation amount, the lubricant deterioration amount, the indentation / lubricant identification amount are specified, and the time frequency distribution (wavelet) is displayed. ,
8. The remaining life diagnosis method for a rolling bearing according to claim 1 , wherein a failure mode is determined and diagnosed by a human. In the determination step, when the result of indentation detection exceeds the threshold value for the diagnostic rolling bearing (3), and the cepstrum cootosis exceeds the threshold value, the indentation origin type peeling mode is set. There is a problem, it is determined that there is the generation of indentation, to assessing the remaining service life in the indentation-originated flaking mode, according to claim 1, 5, 6 or 7 remaining life diagnosis method of a rolling bearing of which is characterized in that . If it is determined in the determination step that the indentation start-type separation mode of the diagnostic rolling bearing (3) is abnormal, or if it is determined that there is an abnormality in the lubricant deterioration mode, it is recognized that attention is required, and the amount of indentation, lubrication 8. The oil deterioration amount, the indentation / lubricant identification amount are specified, the time frequency distribution (wavelet) is displayed, the failure mode is determined and diagnosed by a human, and the diagnosis according to claim 1, 5 , 6, or 7 A method for diagnosing the remaining life of rolling bearings. In the determination step, threshold values for detecting indentation and lubricant deterioration are (a) BRG size, (b) equipment weight, (c) motor output, (d) equipment type, (e) these (a) The remaining life diagnosis method for a rolling bearing according to claim 4 or 14 , wherein the remaining life diagnosis method is changed according to a combination of (d). In the judgment stage, in addition to detection of indentation and lubricant deterioration, all faults of rotating machines are added, including conventional fault detection methods that diagnose equipment unbalance, misalignment, loose foundations, etc. by frequency analysis. The remaining life diagnosis method for a rolling bearing according to any one of claims 1 to 26, characterized by: