RU2104510C1 - Method of complex vibration diagnosis of antifriction bearings and device intended for its realization - Google Patents

Abstract

FIELD: bearing industry. SUBSTANCE: vibration signal of bearing is measured when the latter rotates under load. Vibration signal is converted to digital form and entered into operating buffer in different modes. Values of rotational speed of rotor, of cage with balls relative to external and internal races, and balls around their axis, and peak frequency are determined. Four components of vibration signal depending on quality of bearing member manufacture are single out. Vibration level and power contribution of every component into total power of vibration signal are determined. Condition of bearing members is judged with due regard for known properties of vibration signals incorporated in computer memory. Device intended for realization of this method has vibrator power supply, amplifier-corrector, high- and low-pass filters, analog-to-digital converter. Furrier converter, meter of rotational speed and periods, separator of linear and continuous spectra, peak frequency meter, formers of vibration signals of external and internal races, balls and contamination. Vibration signal formers are connected to display and printer. EFFECT: more efficient diagnosis. 3 cl, 1 tbl, 1 dwg

Description

 The invention relates to the bearing industry and can be mainly used to control finished products in order to determine the quality of production by analyzing the vibrational characteristics of the bearing.

 A known method for the diagnosis of rolling bearings, which is based on monitoring changes in the root mean square level and spectral density of vibration signals / cm. article "Detection of damage to the rolling bearing by statistical analysis of vibration", the magazine "Design", 1978, t. 100, N 2 /.

 The disadvantages of the method is the inability to determine the quality of manufacture of all its parts separately, as well as the degree of contamination.

 A known device for the diagnosis of rolling bearings, in which the method of analysis of vibration signals / cm is used. author testimonial. USSR 540186, class G 01 M 13/04, 1977 /. It contains a vibration converter, an amplifier, a frequency analyzer, a synchronizer, a counter, a matrix decoder, a coincidence unit, a mode switch, a defect probability input unit, a storage unit, an adder, a threshold unit, and a digital converter-amplifier. However, such a device is complex in design and allows you to determine only fairly rough local defects of the bearing parts, resulting from mechanical impulses (shock), periodically repeating with frequencies that are called bearing frequencies.

 There is also known a method and apparatus for assessing non-ideal surface geometry of rings or balls by analysis of vibration signals, / see journal "Design and technology of mechanical engineering", 1987, N 1, p. 140 - 149 /. The method is based on the analysis of a short vibrational signal excited when balls with defects meet on a rotating ring or a rotating ring with balls defects in an assembled bearing. The device contains a device that reproduces vibrational signals, a high-pass filter, a digital tape recorder, a sensor, a speaker, a low-pass filter, an amplifier, an analog-to-digital converter, and a computer.

 The disadvantages of the device and method are the low technological capabilities in the diagnosis of bearings, the low accuracy of the microgeometry of the surfaces of rings and balls, as well as the low productivity of the control process, since the probe can be checked only after disassembling the bearing.

 It is also known a device for vibration diagnostics of rolling bearings, equipped with two analog-to-digital converters, an interface, a printing device, a display, as well as a synthesizer and two parallel channels of the cosine and sine components of the vibration signal, / cm author testimonial. USSR 1620881, class G 01 M 13/04, 1991 /.

 However, this method of diagnosis and the device does not allow you to quickly and accurately conduct comprehensive vibration diagnostics of all parts of the bearing without disassembling it to determine the quality of their manufacture, as well as assessing the degree of contamination of the bearing.

 The technical result of the invention is to provide without disassembling the bearing estimates with a high degree of accuracy of the quality of manufacture of the bearing by taking into account the phase changes in the vibration signal over time, as well as the degree of contamination.

 This is achieved by the fact that in the method of complex vibration diagnostics of rolling bearings, the vibration signal is measured when it is rotated under load, the vibration signal is sampled and digitized and introduced into the working buffer in different modes, changing the sampling frequency, filter cutoff frequency and input time, in each of the modes select the optimal gain, in each of the modes the vibration signal is subjected to a fast Fourier transform, the individual rotor speeds of the rotor, the separator with ricks relative to the outer and inner rings, the ball around its own axis and the peak frequency, using the frequency of the vibration signal associated with the relative frequencies of rotation of the bearing parts, and isolate the four components of these signals, due to the manufacturing quality of the inner and outer rings, balls and the degree of contamination, for each from these components of the signal determine the level of vibration and the contribution of power to the vibration of each of the components in the total power of the vibration signal and, using alozhennye in computer memory known properties of vibration signals, judging the state of the elements. The vibration signal can be entered in three modes using the following parameters: sampling frequency - 800 Hz, 5000 Hz, 25000 Hz, cutoff frequency - 300 LF, Hz, 2000 LF, Hz, 10000 LF, Hz, input time: first mode - 40 s , second 10 s, third - 2 s.

 A device for comprehensive vibration diagnostics of bearings includes a vibration transducer, an amplifier-corrector, a high-pass and low-pass filter, an analog-to-digital converter, a Fourier transducer, a determinant of frequency and periods of rotation, a line and continuous spectrum separator, a peak frequency counter, vibration shapers of the outer ring, inner ring, balls and dirt, the output of the latter is connected to the display and the printing device, while the second output and input of the determinant of frequency and rotation periods Nia frequency counter connected to the peak frequency, and outputs the second input of the vibration signals of the outer ring and the inner and balls are connected to the block inclusion and exclusion peaks, and outputs a third input of the outer ring and inner ring are connected to the frequency setting elements of these rings.

 The drawing shows a diagram of a device.

 The device comprises a vibration transducer 1 mounted on a test bearing mounted on a rotating rotor. An amplifier-corrector 2 is connected to its output, to which a filter 3, a block 4 for preparing a vibration signal, input and ADC, a Fourier converter 5, a determinant of 6 frequencies and periods of rotation by time periods, a separator 8 of line and continuous spectra are connected in series. The second output and input of the determinant 6 frequencies are connected to the frequency meter 7 peak frequencies. The output of the spectral splitter 8 is connected in series to the outer ring vibrator 9, the inner ring vibrator 11, the balls 13 vibrator 13 and the dirt vibrator 15, and the output of the latter through the AOC 16 and interface 17 is connected to the display and the printing device. The second outputs and inputs of the shapers of vibration signals 9, 11, 13, respectively, of the outer ring, inner ring and balls are connected to the block 14 on-exclusion of peaks. The third output and input of the outer ring vibration generator 9 is connected to the outer ring frequency setter 10, and the third output and input of the inner ring vibration generator 11 is connected to the inner ring frequency setter 12.

 The device operates as follows.

 The mechanical vibrations of the bearing, which appear due to inaccuracies in manufacturing and assembly, are detected by the vibration transducer 1, which converts them into a proportional electrical signal. From the amplifier-corrector 2, designed to match the high output impedance of the vibration transducer 1 with subsequent elements and amplification, the vibration signal is fed to the input of the bandpass filter 3 and through it to the unit 4 for preparing and inputting the vibration signal. Input is carried out in three modes. Their parameters are given in table 1.

 In each of these modes, the optimal gain is first selected, i.e. the maximum coefficient at which the ADC 16 does not overflow, and then data is already entered into the working buffer.

 The frequency of rotation of the inner ring and the set of balls relative to the inner and outer rings is determined from the incoming vibration signal using the investigated properties of the vibration signals.

 Initially, the vibration signal recorded in the first mode is subjected to fast Fourier transform using 8192 points with averaging rms values over the vibration signals. As a result, the bearing receives vibration signals in the range up to 400 Hz with a frequency resolution of Δff = 0.0977 Hz. Since when the signal was input, the cut-off frequency of the low-pass filter was 300 Hz, then practically the significant components of the vibration signal are received at a frequency not higher than the cut-off frequency.

,
Она определяется таким образом с точностью лучше, чем 0,05 Гц. Такая точность необходима, чтобы различать гармоники, вплоть до конца диапазона исследования вибрации подшипника, т. е. до 10000 Гц. При такой точности ошибка, которая набегает для последней гармоники, (имеющей номер 10000/25 = 40), составит не более 40•0,05 = 2 Гц.In this vibration signal there is a significant component corresponding to the rotational speed of the inner ring. It is known approximately from the rotation mode of the bearing: slightly less than 25 Hz. There are no other components of the vibration signal nearby. This is the basis for the algorithm for accurately determining the rotation frequency of the inner ring. First, determine the component of the vibration signal in the range of 24 - 25 Hz, having a maximum amplitude. Then take the two previous components. The rotational speed is defined as the weighted average of the frequencies of these five components, where the weights are the squares of the corresponding amplitudes. If we denote the maximum harmonic number by f _BP , the frequencies of the five initial harmonics are f _i-2 , f _i-1 , f _i , f _{i + 2} , f _{i + 2} , and the amplitudes a _i-2 , a _i-1 , a _i , a _{i + 1} , a _{i + 2} , then the speed value is determined by the formula (1):

,
It is determined in this way with an accuracy better than 0.05 Hz. Such accuracy is necessary to distinguish harmonics, up to the end of the range of investigation of bearing vibration, i.e., up to 10,000 Hz. With such accuracy, the error that occurs for the last harmonic (having the number 10000/25 = 40) will be no more than 40 • 0.05 = 2 Hz.

 Next, the same speed determines the speed of the set of balls relative to the outer ring. The harmonic corresponding to this frequency is usually very weakly expressed, and often completely invisible against the general background. At the same time, the harmonic corresponding to the frequency of passage of the balls along the outer ring, which is equal to the desired frequency times the number of balls in the bearing, is pronounced. Knowing approximately the frequency of rotation of the set of balls relative to the outer ring (9.3 - 10 Hz), multiplying the boundaries of this range by the number of balls gives a new range in which, acting according to the algorithm described for the frequency of rotation of the inner ring (i.e., determining at the beginning the maximum component in the obtained range and ending with the calculation according to formula (1), determine the frequency of passage of the balls along the outer ring with an accuracy of no worse than 0.05 Hz. Dividing this frequency by the number of balls, we obtain the rotation speed of the set of balls relative relative to the outer ring with an accuracy of at least 0.01 Hz, since the number of balls in the bearing is at least 6.

After determining the rotational speed of the outer ring (f _p ) and the rotational speed of the set of balls relative to the outer ring (f _n ), calculate the rotational speed of the set of balls relative to the inner ring using the formula
f _in = f _p - f _n (2)
On the basis of certain three of the four frequencies, the ruled part of the vibration spectrum of the bearing is built. The fourth frequency of rotation of the ball around its own axis (f _W ) is determined later.

 Three sets of harmonics can be distinguished in the ruled part of the ball bearing vibration signal, which are excited respectively by the outer and inner rings and a set of balls. Let n be the number of balls in the bearing. Let k denote an integer whose variation range will be indicated separately for each harmonic group. Let m be another integer in the range from 0 to 5.

Then the harmonic frequencies corresponding to the excitation from the outer ring will be
k • n • f _n ± m • f _p , (3)
where k = 1 - 125.

Harmonic frequencies corresponding to excitation from the inner ring
k • n • f _at ± m • f _p , (4)
where k = 1 - 85.

Harmonic frequencies corresponding to excitation from a set of balls:
m • f _n (5)
k • f _w ± m • f _n , (6)
where k = 1 - 200.

 All of these components will not necessarily be represented in fact in a real vibration signal. So, for example, for the outer ring in formula (3) there are mainly harmonics corresponding to m = 0. For the inner ring, on the contrary, harmonics corresponding to m = 0 in formula (4) rarely occur, and side frequencies are present. The same is true for balls: harmonics corresponding to m = 0 in formula (6) are usually weakly expressed, and lateral frequencies are present. The frequencies described by formula (5) and characterizing the different sizes of the balls do not always appear. However, the diagnostic program each time individually for each bearing analyzes the presence and absence of all these harmonics and determines their amplitude.

 This structure of the vibration signal is confirmed by the results of detailed experimental and theoretical studies and is fully consistent with the actual vibration signal of the rolling bearing. Therefore, it is the basis for constructing an adequate system of vibration diagnostics of ball bearings.

 The formation of the vibration signal is as follows. First, all peaks in the vibration signal are distinguished. For this, the moving average vibration signal is calculated in the range of 20 previous and 20 subsequent harmonics. Then, the harmonics of the initial vibration signal are examined sequentially, and if its amplitude exceeds the corresponding average by a factor of two, it is considered an element of the peak. Peak elements traveling consecutively without gaps are considered to be a single peak element. To identify individual peaks in the process of enumerating by the adjuster 10 and 12, the numbers of harmonics that are the first and last elements of each of the peaks are recorded in a special array of pairs of numbers. From the obtained set of peaks, those that correspond to formulas (3) and (4) in the frequency range under consideration are selected. They are simultaneously excluded from the initial spectrum and included in the vibration spectrum of the outer ring. The same operation is repeated for the inner ring using the frequencies described by formula (5). Then only peaks corresponding to balls remain in the initial spectrum. These peaks are excluded from the initial spectrum and are included in the spectrum of vibration excited by the balls.

 The result is four vibration signals corresponding to excitation from four main causes: the outer ring, inner ring, balls and contamination. The last spectrum is obtained after the exclusion of all peaks corresponding to the ruled part of the vibration signal.

 In parallel with the formation of vibration signals, a quadratic summation of the components of each of the signals is performed. Moreover, for each of the signals two sums are calculated. One determines the acceleration of vibration, and the other determines speed. For speed, the component of the vibration signal is previously divided by the circular frequency of this harmonic. The roots are square of the eight values obtained and are the vibration level for acceleration and speed in the frequency range. This data is converted to decibels.

 Then carry out the diagnosis in two other frequency ranges. For the range of 300 Hz - 1800 Hz, the data entered in mode 2 are used, and for the range of 1800 Hz - 10,000 Hz - the data entered in mode 3 (see Table 1). To obtain the initial vibration signal, first, the fast Fourier transform, then carry out the remaining operations described above.

 The results of comprehensive vibration diagnostics of a rolling bearing are obtained in the form of vibration signals excited by four factors in three frequency ranges and vibration levels in these ranges for speed and acceleration. The vibration levels of speed are displayed on the display screen 17 and at the same time on the printing device 18. If necessary, you can display the vibration levels for acceleration. Vibration signals can also be displayed and printed if necessary.

Claims (3)

 1. The method of complex vibration diagnostics of rolling bearings, which consists in measuring the vibration signal of the bearing when it rotates under load, the vibration signal is sampled and converted into digital form and entered into the working buffer in different modes, changing the filter cutoff frequency and input time, in each of the modes select the optimal gain, in each of the modes the vibration signal is subjected to a fast Fourier transform, the individual rotor speeds of the rotor, the cage with balls are determined individually for each bearing the outer and inner rings, the ball around its own axis and the peak frequency, using the frequency of the vibration signal associated with the relative frequencies of rotation of the bearing parts, and four components of these signals are determined, due to the manufacturing quality of the inner and outer rings, balls and the degree of contamination, for each of these components of the signal determine the level of vibration and the fractional contribution of power to the vibration of each of the components in the total power of the vibration signal and, using computer memory known properties of vibration signals, judge the state of the elements.  2. The method according to claim 1, characterized in that the vibration signal is introduced in three modes using the following parameters: sampling frequency 800, 5000, 25000 Hz, cutoff frequency 300, 2000, 10000 Nh / Hz, input time first mode 40 s, second 10 s, third 2 s.  3. A device for complex vibration diagnostics of rolling bearings, containing a vibration transducer, an amplifier-corrector, a high and low frequency filter, an analog-to-digital converter, a Fourier converter, a determinant of frequency and periods of rotation, a line and continuous spectrum separator, a peak frequency counter, a peak frequency counter, an external ring vibration conditioner , inner ring, balls and dirt, the output of the latter is connected to the display and the printing device, while the second output and input of the frequency determiner and IRS rotation frequency counter connected to the peak frequency, and outputs the second input of the vibration signals of the outer ring and inner rings and balls are connected to the block inclusion and exclusion peaks, and third inputs and outputs vibration signals formers the outer ring and the inner Colzano connected to the frequency setting elements of these rings.

Images