DE102011055523A1 - Method for vibration analysis of e.g. bearing element, involves interpreting measurement values, which exceed predetermined threshold value, as oscillation pulse, and processing measurement values by algorithm - Google Patents

Abstract

The method involves producing measurement values by a sensor device (15) due to oscillations, where the measurement values are supplied to an evaluating device (20) as input values. The measurement values are processed by an algorithm (21) stored in the evaluating device. An erroneous operation of a mechanical equipment (10) or faulty machine elements is stopped while exceeding threshold values stored in the evaluating device with respect to processing. The measurement values, which exceed a predetermined threshold value, are interpreted as oscillation pulse and processed by the algorithm. The sensor device is designed as an acceleration sensor, an impact sound sensor, a microphone, an acoustic emission sensor, an ultrasonic sensor or a hydrophone sensor. An independent claim is also included for a system for vibration analysis of a mechanical equipment or machine elements.

Description

The invention relates to a method for vibration analysis of equipment or machine elements, in particular of rotating machine elements such as bearing elements or shafts, according to the preamble of claim 1. Furthermore, the invention relates to an apparatus for performing a method according to the invention.

A method according to the preamble of claim 1 is already known from practice and serves to detect and analyze vibrations occurring during the operation of equipment or machine elements due to, for example, wear or other events occurring. At least one sensor device is used for this purpose, which generates measurement signals on the basis of the vibrations, which signals are fed to an evaluation device as input values. In the evaluation device, at least one algorithm is provided which processes the measured values in order, for example, to conclude on the occurrence of periodically occurring vibration signals on a faulty bearing element or the like. In particular, if the evaluation of the vibration signals over a longer period of time or continuously, it can be concluded on emerging damage to the system or the machine element. For this purpose, the sensor device used to detect the oscillation signals receives the measured values or oscillation signals regularly or in certain time periods regularly and forwards them to the evaluation device. In the evaluation device or an upstream device, the measured values, which may be recorded analogously, are usually digitized in order to be able to process them in the evaluation device.

A disadvantage of the method known from the prior art is that all measured values transmitted from the sensor device to the evaluation device are processed by means of the algorithm. This means in practice that the data processing systems used for processing by means of the evaluation algorithm require a relatively high storage capacity and a relatively high computer power in order to be able to process the total number of data or measured values supplied by the sensor device to the evaluation device as input values. In addition, the frequency required in a continuous detection of the vibration signals at a digitization of the vibration signals is at least twice the maximum frequency of the vibration signals in order to interpret the vibration signal without error as such. In addition, monitoring the condition of the equipment often requires measuring multi-channel vibration signals as a result of further increasing the number of measurements. With such measurements, therefore, additional requirements are added to the measurement data management, the computer performance and the storage capacity. In particular, the latter situation applies to sensor devices which comprise ultrasonic sensors. These lead due to high frequency components to very high sampling frequencies and consequently to very large required storage capacities and computer power. The sampling frequencies may be several hertz, wherein the ultrasonic signals are subdivided the vibration signals.

From the prior art, the so-called envelope curve analysis is also known, which is used for the demodulation of vibration signals. Envelope analysis is performed by extracting the envelope (the vibration pulses) followed by Fourier transformation. Envelope analysis is useful for detecting periodic events in vibration signals. As a result, changes that affect an intended function of the system or of the machine element can be detected.

In view of the above-mentioned prior art, the inventor has set itself the task of developing a method for vibration analysis of equipment or machine elements according to the preamble of claim 1 such that they are compared to the prior art with data processing systems with lower computer power or storage capacity can. This object is achieved in a method according to claim 1, characterized in that only measured values of vibration signals that exceed a certain threshold, interpreted as (relevant) vibration pulse and processed by means of the stored in the evaluation algorithm. In other words, this means that in contrast to the prior art, in which all the vibration signals that are detected by the sensor device, be it continuously or in the sense of a cyclic detection of the vibration signals, are processed in analog or digital form, only quite certain vibration signals, which are based on the height of the measured value assumes that it is a relevant for the analysis or evaluation of the condition of the system or the machine element vibration pulse processed.

Advantageous developments of the method according to the invention for the vibration analysis of equipment or machine elements are listed in the subclaims. In the context of the invention, all combinations of at least two features disclosed in the claims, the description and / or the figures.

For the detection of possibly relevant vibration signals or of measured values that might contain a relevant vibration pulse, it is proposed that the measured values (of the sensor device) be processed before the comparison with the threshold value for increasing the pulse strength. To increase the momentum content of the oscillation signal, preprocessing steps, such. As the use of a high-pass filter, apply. Such processing of the input signals can be done both in analog and in digital form.

Quite preferred is a method, wherein the algorithm comprises a processing of the measured values by a spectral estimation in order to generate a spectrum of the measured values. Such a spectral estimation can be done for example by the Lomb Scargle Periodogram, the Korenberg Method or the Palmer Chi-squared method. The spectral estimation makes it possible to find periodicities in vibration signals. These periodicities may indicate changes in a plant or machine element that are detrimental to its intended proper function. Such periodicities can be caused, for example, by pulse trains as a result of rolling over damaged bearings in a rolling bearing.

Moreover, it is very particularly preferred if an evaluation of the vibration signals takes place based on the generated spectrum. Such an evaluation of the vibration signals can lead to the fact that in the case of corresponding vibration signals, for example, a damage occurring in the near future or damage to the plant can be referred to, so that by issuing a corresponding warning signal to an operator or by a forced shutdown of the plant or similar to such damage can be avoided.

In order to additionally reduce the required computer power as well as storage capacity at the data processing system, it can be provided that the detection of the measured values by the at least one sensor device takes place discontinuously. A discontinuous detection here means either a recording of measured values controlled by a random generator or a recording of the measured values after respectively fixed pauses in which no measured values are recorded.

A further reduction of the computer power and the storage capacity takes place when only that measured value is processed within a vibration pulse at which the measured value of the vibration pulse has its maximum. What is meant here is the fact that within a vibration pulse, a multiplicity of measured values is generated in which the measured values lie above the threshold value according to the invention. In order to prevent all of these measured values which are above the threshold value from having to be processed, it is assumed in the simplest case that only the measured value with respect to its time and its oscillation amplitude is used for the vibration analysis in which the measured value has the maximum.

Alternatively, however, it may also be provided that measured values which are recorded before and / or after the maximum of a maximum measured value (within the oscillation pulse) are taken into account. In this case, it is basically possible to process any number of measured values in which the respective measured value is above the threshold value.

In order to be able to delineate regular oscillation pulses of one-time events, in which there is normally no damage or permanent impairment of the operation of the plant or of the machine element, it is also proposed that only oscillation pulses with a frequency greater than 0.01 Hz be taken into account.

The method according to the invention is characterized in that a multiplicity of different sensor types can be used for the sensor device for detecting the vibration signals. In particular, acceleration sensors, structure-borne sound sensors, microphones, acoustic emission sensors, ultrasonic sensors or hydrophone sensors are intended here.

The invention also includes an apparatus for carrying out a method according to the invention. This device comprises at least one sensor device for detecting a vibration signal, an evaluation device for converting the acquired measured values by means of an algorithm stored in the evaluation device and a data output unit for displaying results. Such a device comprises a comprehensive evaluation and presentation of vibration signals, as they occur during operation of equipment or machine elements, in particular to point out impending damage events or improper operation.

Further advantages, features and details of the invention will become apparent from the following Description of preferred embodiments and with reference to the drawing.

This shows in:

1 a schematic representation of a device for vibration analysis in a system,

2 and 3 simplified representations of measured values acquired by means of a vibration sensor for representing different methods according to the invention and

4 a spectrum formed from measured values by means of the method according to the invention.

In the 1 is part of an arbitrary machinery 10 shown. The machinery 10 includes one in a longitudinal axis 11 rotating shaft 12 , in two axially spaced bearings, for example in rolling bearings 13 . 14 , stored. On the radial outer surface of a rolling bearing 13 is a sensor element 15 arranged the vibrations of the rolling bearing 13 in the direction of the double arrow 16 detected.

In the sensor element 15 it is in particular an acceleration sensor or a structure-borne sound sensor. Depending on the purpose or application of the sensor element 15 However, this can also be designed as a microphone, as an acoustic emission sensor, as an ultrasonic sensor or as a hydrophone sensor.

In a modification of the illustrated embodiment, it may of course also be provided that a second, perpendicular to the first sensor element 15 arranged sensor is used to record also vibrations that are perpendicular to the direction of the double arrow 16 run. Furthermore, it is also within the scope of the invention, more than one sensor element 15 or two sensor elements 15 to use, in particular, each of the two rolling bearings 13 . 14 each with at least one sensor element 15 equip. Also, it is conceivable the wave 12 or their concentricity with a sensor element 15 to monitor.

The sensor element 15 is over a line 17 with an evaluation device 20 coupled. The evaluation device 20 comprises at least one electronic data processing system with a computer and means for at least temporary storage of measured values and one by the reference numeral 21 characterized algorithm for processing the from the sensor element 15 the evaluation device 20 Measured values supplied as input signals. The evaluation device 20 is over another line 22 with an input / output unit 25 coupled. The input / output unit 25 For example, it may include a keyboard (not shown) for entering data and a screen for displaying results. Furthermore, it can be provided that the evaluation device 20 and / or the input / output unit 25 with the control of the machinery 10 coupled to the occurrence of relevant (damage) events in the operation of the machinery 10 intervene without the need for an operator.

From the sensor element 15 Measured values are recorded continuously or discontinuously due to vibrations of the rolling bearing 13 in the direction of the double arrow 16 be generated. These measured values are the evaluation device 20 fed. Before the readings through the algorithm 21 the evaluation device 20 be evaluated or processed, it may be provided that the measured values before processing by means of the algorithm 21 be edited to increase the impulsiveness, for example by means of a high-pass filter. In this case, both an analog and a digital preprocessing of the measurement signals is possible.

In the evaluation device 20 or in the algorithm 21 According to the invention, it is provided that at least one threshold value S is stored, which, when exceeded, can be used to conclude a relevant oscillation pulse. In this case, the threshold value S, relative to the direction of the oscillation pulse or of the measured value, can be in both the positive and the negative direction (with reference to the double arrow 16 ).

The threshold value S may be a constant threshold value. However, it is also conceivable that the threshold value S is a dynamic threshold value S which is adapted in particular on the basis of the signal profile or the measurement signals over time.

In the 2 are those of the sensor element 1 recorded measured values or vibration amplitudes A over time t by way of example and to simplify the explanation of the invention. Here, the case is assumed that the (positive) threshold S within the illustrated period in two areas 27 . 28 is exceeded. The first area 27 begins at time t ₁ and ends at time t ₂ . The first area 27 has its maximum _Ami at time t _m1 . The second area 28 begins at time t ₃ and ends at time t ₄ . He reaches his maximum office at the time t _m2 .

By way of example, in a first embodiment of the invention, only the maxima A _m1 and A _{m2 of} the vibration signals are determined by the algorithm 21 processed. This means that the evaluation device 20 or the algorithm 21 only has to be designed or designed to process two measured values A _m1 and A _m2 at two times t _m1 and t _m2 .

In a modification of the embodiment described above, it can also be provided that the measured values S at the times t ₁ to t _{4 are} likewise determined by the algorithm 21 to be edited.

In the 3 the same vibration signals are shown as in the 2 , In amendment to 2 however, are in each of the two areas 27 and 28 in each case two further points in time t ₅ and t ₆ as well as t ₇ and t _{8 are} designated, in which the measured values of the sensor element 15 By way of example, 80% of the respective maximum value A _m1 and A _m2 reach, if this value is above the threshold value S. In this embodiment, it is assumed that both the measured values at the times t _m1 and t _m2 and the measured values at the times t ₅ and t ₆ as well as t ₇ and t _{8 are determined} by means of the algorithm 21 are processed.

It is of course within the scope of the invention, in amendment of 2 and 3 any number of measured values through the algorithm 21 if these measured values are above the threshold value S. In the context of the invention, it is also possible to use other criteria with regard to relevant measured values. For example, measured values that occur only in an unperiodic or relatively large time interval can be ignored. However, the sampling rate or measuring frequency is preferably more than 0.1 Hz.

The by means of the algorithm 21 in the evaluation device 20 processed measured values or oscillation signals are interpreted or processed as the smallest possible information for vibration analysis. Of course, all temporal manipulations are included in irregular signals, such as those used in order analyzes. These manipulations can lead to an improvement in the evaluation of the measurement signals.

Due to the measured values generated by the algorithm 21 carried out a spectral estimation of irregularly occurring vibration signals. This spectral estimation makes it possible to find periodicities in the vibration signal.

In the 4 is the result of such an estimate. One recognizes in particular two peaks 29 . 30 in which an amplitude of 2 or 3 is achieved, and wherein the maximum is at a frequency of about 8 Hz or 19 Hz. Because of these characteristic maxima can by means of the evaluation 20 on a certain damage pattern or a certain vibration condition of the machinery 10 or the rolling bearing 13 be closed, what about the input / output unit 25 the results are displayed and, if necessary, countermeasures that can be taken via the machine control can be taken.

The procedure described so far for the vibration analysis of machinery 10 or machine elements can be modified or modified in many ways, without departing from the spirit. This consists in the selection of measurement data to reduce the required computer power or storage capacity in an evaluation 20 by only those measured values of a sensor element 15 be considered, in which a threshold value S is exceeded.

LIST OF REFERENCE NUMBERS

10

machinery

11

longitudinal axis

12

wave

13

roller bearing

14

roller bearing

15

sensor element

16

double arrow

17

management

20

evaluation

21

algorithm

22

management

25

Input / output unit

27

first area

28

second area

29

peak

30

peak

S

threshold

t

Time

t ₁

Time (start - first range)

t ₂

Time (end - first range)

t _m1

Time (maximum - first range)

t ₅

further time (first area)

t ₆

further time (first area)

t ₃

Time (beginning - second area)

t ₄

Time (end - second range)

t _m2

Time (maximum - second range)

t ₇

further time (second area)

t ₈

further time (second area)

Claims (10)

Method for vibration analysis of installations ( 10 ) or machine elements, in particular of rotating machine elements such as bearing elements ( 13 . 14 ) or waves ( 12 ), in which by means of at least one sensor device ( 15 ) are generated due to vibrations measured values (A), the an evaluation device ( 20 ) are supplied as input values, that the measured values (A) by means of at least one in the evaluation device ( 20 ) stored algorithm ( 21 ) and that when exceeding in the evaluation device ( 20 ) with regard to the results of the processing of stored limit values on faulty operation of the plant ( 10 ) or a faulty machine element is closed, characterized in that only those measured values (A) which exceed a specific threshold value (S) are interpreted as a vibration pulse and by means of the in the evaluation device ( 20 ) stored algorithm ( 21 ) are processed. A method according to claim 1, characterized in that the measured values (A) are processed before the comparison with the threshold value (S) for increasing the pulse content. Method according to claim 1 or 2, characterized in that the algorithm ( 21 ) comprises processing the measured values (A) by means of a spectral estimation in order to generate a spectrum of the measured values (A). A method according to claim 3, characterized in that based on the generated spectrum, an evaluation of the vibration signals. Method according to one of claims 1 to 4, characterized in that the detection of the measured values (A) by the at least one sensor device ( 15 ) is discontinuous. Method according to one of claims 1 to 5, characterized in that only the measured value (A _m1 , A _m2 ) is processed within an oscillation pulse at which the measured value (A) of the oscillation pulse has its maximum. Method according to one of claims 1 to 5, characterized in that measured values (A), which are detected before and / or after the maximum of a vibration pulse, are taken into account. Method according to one of claims 1 to 7, characterized in that only vibration pulses with a frequency greater than 0.01 Hz are taken into account. Method according to claim 8, characterized in that as sensor device ( 15 ) an acceleration sensor, a structure-borne sound sensor, a microphone, an acoustic emission sensor, an ultrasonic sensor or a hydrophone sensor is used. Device for carrying out a method according to one of claims 1 to 9, comprising at least one sensor device ( 15 ) for detecting a vibration signal, an evaluation device ( 20 ) for converting the acquired measured values (A) by means of one in the evaluation device ( 20 ) stored algorithm ( 21 ) and a data output unit ( 25 ) for the presentation of results.

Images