A method for detecting the position of a rotating machinery collision and impact fault

Abstract

The present invention discloses a method for detecting the fault position of a rotating machine due to friction impact, and the fault position is determined by waveform reconstruction and correlation analysis. The method comprises the following steps: step 10) installing a detection component; step 20) extracting a signal from the rotating machine to obtain a time domain signal when an abnormal friction impact occurs; step 30) reconstructing the time domain signal; step 40) analyzing the signal to obtain the time difference experienced by the signal being transmitted to the left and right speed sensors; step 50) fault location: assuming that the propagation speed of the mechanical wave in the rotating shaft is V, and the distance between the speed sensors at the two ends of the rotating shaft is S, the position where the friction impact occurs is calculated based on the time difference. The method determines the fault position by waveform reconstruction and correlation analysis.

Description

Technical Field

[0001] The invention belongs to the technical field of power machinery, and in particular relates to a method for detecting the position of a rotating machinery collision and impact fault. Background Art

[0002] The dynamic equation is a non-homogeneous differential equation, and its solution is the linear superposition of the general solution of the corresponding homogeneous differential equation and the special solution that satisfies the conditions. The vibration of the rotor driven by the external environment includes two aspects: one is the intrinsic vibration of the system (vibration at its own natural frequency), and the other is the vibration that follows the frequency of the external force. The intrinsic motion decays exponentially with time, so in steady state the rotor vibrates following the driving frequency, which is forced vibration. When a fault such as severe rotor friction occurs, the vibration spectrum generated by the force impact effect is very rich. In addition to the forced vibration frequency of the system at this time, it also contains a large number of low-frequency and high-frequency components, and at the same time it will stimulate the free vibration response of the rotor, that is, the vibration signal will contain the rotor's natural frequency component.

[0003] With the development of large-scale rotating machinery, the dynamic and static clearance is usually designed to be relatively small. During operation, dynamic and static friction may occur if you are not careful. In mild cases, it will lead to a decrease in the efficiency of the rotating machinery, and in severe cases, it will cause large vibrations, affecting the safety of operation. After the rotor friction occurs, due to the high linear speed of the rotating shaft surface of the rotating machinery, friction will generate high temperature. The temperature of the area with severe friction is higher than that of the area with mild friction and no friction. The uneven temperature will cause thermal deformation of the rotor, that is, the force impact effect of the friction brings about a thermal shock effect. The most direct consequence of the thermal shock effect is that it causes thermal deformation of the shaft, increases the unbalanced force, and increases the vibration of the rotating machinery. Large vibrations will further aggravate the degree of friction, thus forming a vicious circle, which will cause great harm to the safe operation of the rotating machinery. Therefore, it is necessary to pay attention to the vibration caused by force impact effects such as friction.

[0004] The existing monitoring of friction fault sites mainly relies on the vector monitoring method. To monitor the vibration caused by friction, many factors must be taken into account, including the amplitude of the vibration, the amount of change in the vibration, the phase change of the vibration, etc. The comprehensive consideration of the vector monitoring method is more effective. When friction occurs on the surface of the shaft, the temporary thermal effect will cause the imbalance of the shaft to change, and the synthetic vibration vector will also change accordingly. The friction position can be found by using the change relationship of this vibration vector. Select a suitable position on the rotor to install the displacement sensor, perform a fast FFT on the collected frequency-passing signal after pre-amplification and digital filtering, extract the industrial frequency component of the vibration signal, and then compare and calculate the industrial frequency component with the industrial frequency vibration component retrieved from the historical database to obtain the change in the vibration vector caused by the friction, thereby determining the location of the friction fault.

[0005] Disadvantages of existing technology:

[0006] (1) The vector monitoring method uses the change in unbalance caused by the thermal shock effect of friction, and calculates the friction impact as an unbalanced force. The friction position is determined by comparing the vibration vector changes at both ends of the shaft. This method is affected by the unbalanced force of the rotor itself. At the same time, in most cases, the vibration vector change is more obvious only near the friction point. The farther away from the friction point, the smaller the vibration vector change.

[0007] (2) An important step in the vector monitoring method is to obtain information from the historical database for comparison. When the historical operation data is insufficient or there are certain errors in the reference vibration vector value in the historical data, it will affect the judgment of the results. Summary of the invention

[0008] Technical problem: The present invention proposes a method for detecting the fault location of a rotating machine due to friction and impact, and determines the fault location through waveform reconstruction and correlation analysis.

[0009] Technical solution: To solve the above technical problems, the embodiment of the present invention adopts the following technical solution:

[0010] A method for detecting the position of a rotating machinery collision and impact fault comprises the following steps:

[0011] Step 10) installing the detection component;

[0012] Step 20) When an abnormal friction impact occurs, extract a signal from the rotating machine to obtain a time domain signal;

[0013] Step 30) reconstructing the time domain signal;

[0014] Step 40) Analyze the signal to obtain the time difference Δt experienced by the signal being transmitted to the left and right speed sensors;

[0015] Step 50) Fault location: Assuming that the propagation speed of the mechanical wave in the rotating shaft (1) is V, and the distance between the speed sensors at both ends of the rotating shaft (1) is S, according to the time difference Δt, the position x where the friction impact occurs is as shown in formula (4):

[0016] x=S / 2-VΔt / 2 Formula (4)

[0017] In formula (3), the position x where the friction impact occurs is the speed sensor on the left side of the shaft as the coordinate zero point.

[0018] As a preferred example, the step 10) specifically comprises: installing a speed sensor on each of the bearing seats at the left and right ends of the rotor shaft (1) of the rotating machine to be tested.

[0019] As a preferred example, in the step 10), when the speed sensor is installed, the fault location is between the locations of two speed sensors.

[0020] As a preferred example, the step 20) specifically includes: when an abnormal friction impact occurs on the rotor shaft (1), the signals measured by the two speed sensors are respectively input into a vibration signal acquisition and analysis instrument (8), and the signal acquisition and analysis instrument (8) is used to record the time domain signal after the friction impact occurs.

[0021] As a preferred example, the step 30) specifically includes: the time domain signal includes multiple single waveforms with different amplitudes, phases, and frequencies, as shown in formula (1):

[0022]

[0023] Where y(t) is the vibration displacement of the time domain signal, A1 is the amplitude of the first time domain signal, A2 is the amplitude of the second time domain signal, and A i is the amplitude of the i-th time domain signal, ω1 is the frequency of the first time domain signal, ω2 is the frequency of the second time domain signal, ω i is the frequency of the i-th time domain signal, is the phase angle of the first time domain signal, is the phase angle of the second time domain signal, is the phase angle of the i-th time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, The phase of the sinusoidal signal is damped;

[0024] According to the frequency characteristics of the force impact of the friction, the signals collected by the two velocity sensors are screened and reconstructed respectively. The reconstruction process is as follows: first, the time domain signal is converted into the corresponding frequency domain signal H(f); the upper limit ωmax and lower limit ωmin of the frequency reconstruction are set in the frequency domain signal H(f) to satisfy the condition ωmin<ω0<ωmax, and the other signal frequencies ω1, ω2, ω i are all outside the interval [ωmin, ωmax]; after conversion, the frequencies ω1, ω2, ω i The signal strength of the corresponding signal is represented by A1, A2, A i Convert to B1, B2, B in the frequency domain representation H(f) i , and then the signal strengths B1, B2, B iIt is set to 0, and the frequency domain signal strength corresponding to the interval [ωmin, ωmax] where ω0 is located after the conversion is not changed, so that H(f) can be reconstructed into a new frequency domain signal h(f), in which only the signal corresponding to ω0 exists. Finally, h(f) is inversely transformed into the reconstructed time domain signal.

[0025] After waveform reconstruction, only the time domain waveform caused by abnormal friction impact is retained, as shown in formula (2):

[0026] ζ(t)=ζ0e -βt sin(2π×ω0t) Formula (2)

[0027] Where ζ(t) is the vibration displacement of the reconstructed time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, is the phase of the damped sinusoidal signal.

[0028] As a preferred example, the step 40) includes: reconstructing the two speed sensor signals y1(t) and y2(t) collected by the signal acquisition analyzer (8) in step 30) to obtain the first time domain signal ζ1(t) and the second time domain signal ζ2(t), and obtaining the analysis function R(n) by using the correlation analysis algorithm, as shown in formula (3):

[0029]

[0030] Wherein, R(n) is the cross-correlation analysis function, N is the total number of moments, m represents the moment, and n represents the interval between two signals; ζ1(m) represents the first time domain signal of a speed sensor at moment m; ζ2(m+n) represents the second time domain signal of another speed sensor at moment m+n; when the function R(n) takes the maximum value, the corresponding time interval n=Δt is obtained, and the time difference experienced by the signal to be transmitted to the left and right speed sensors is Δt.

[0031] Beneficial effects: The present invention realizes fault location on the rotating shaft by using two velocity sensors to receive signals at the left and right ends of the friction fault position. According to the propagation speed of mechanical waves in solids and the calculated time difference, the deviation between the position where the friction impact occurs and the center point of the two velocity sensors can be calculated, so that the position where the friction impact occurs can be determined. This method uses the natural frequency characteristics excited by the force impact effect when the friction occurs to reconstruct the waveform and perform correlation analysis on the collected signals, and determines the friction position through local characteristics. There is no need to compare with historical data during the calculation process of the present invention, which is simple to operate and the results are reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1Schematic diagram of a detection device used in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments.

[0034] like Figure 1 As shown, a method for detecting the position of a rotating machinery friction impact fault according to an embodiment of the present invention comprises the following steps:

[0035] Step 10) Install the detection component.

[0036] Step 10) specifically includes: installing a speed sensor on the bearing seats at the left and right ends of the rotor shaft 1 of the rotating machine to be tested. Using two speed sensors to measure the fault position. The speed sensor can be a vibration speed sensor. After the speed sensor is installed, the fault position is between the two speed sensor positions, that is, on the right side of the left sensor and also on the left side of the right sensor.

[0037] Step 20) When an abnormal friction impact occurs, a signal is extracted from the rotating machinery to obtain a time domain signal.

[0038] Step 20) specifically includes: when the rotor shaft 1 has an abnormal friction impact, the signals measured by the two speed sensors are respectively input into the vibration signal acquisition and analysis instrument 8, and the signal acquisition and analysis instrument 8 is used to record the time domain signal after the friction impact occurs.

[0039] Step 30) Reconstruct the time domain signal.

[0040] Step 30) specifically includes: the time domain signal contains multiple single waveforms with different amplitudes, phases, and frequencies, as shown in formula (1):

[0041]

[0042] Where y(t) is the vibration displacement of the time domain signal, A1 is the amplitude of the first time domain signal, A2 is the amplitude of the second time domain signal, and A i is the amplitude of the i-th time domain signal, ω1 is the frequency of the first time domain signal, ω2 is the frequency of the second time domain signal, ω i is the frequency of the i-th time domain signal, is the phase angle of the first time domain signal, is the phase angle of the second time domain signal, is the phase angle of the i-th time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, The phase of the sinusoidal signal is damped;

[0043] According to the frequency characteristics of the force impact of the friction, the signals collected by the two velocity sensors are screened and reconstructed respectively. The reconstruction process is as follows: first, the time domain signal is converted into the corresponding frequency domain signal H(f); the upper limit ωmax and lower limit ωmin of the frequency reconstruction are set in the frequency domain signal H(f) to satisfy the condition ωmin<ω0<ωmax, and the other signal frequencies ω1, ω2, ω i are all outside the interval [ωmin, ωmax]; after conversion, the frequencies ω1, ω2, ω i The signal strength of the corresponding signal is represented by A1, A2, A i Convert to B1, B2, B in the frequency domain representation H(f) i , and then the signal strengths B1, B2, B i It is set to 0, and the frequency domain signal strength corresponding to the interval [ωmin, ωmax] where ω0 is located after the conversion is not changed, so that H(f) can be reconstructed into a new frequency domain signal h(f), in which only the signal corresponding to ω0 exists. Finally, h(f) is inversely transformed into the reconstructed time domain signal.

[0044] After waveform reconstruction, only the time domain waveform caused by abnormal friction impact is retained, as shown in formula (2):

[0045] ζ(t)=ζ0e -βt sin(2π×ω0t) Formula (2)

[0046] Where ζ(t) is the vibration displacement of the reconstructed time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, is the phase of the damped sinusoidal signal.

[0047] Step 40) Analyze the signal to obtain the time difference Δt that the signal takes to be transmitted to the left and right speed sensors.

[0048] The step 40) includes: reconstructing the two speed sensor signals y1(t) and y2(t) collected by the signal acquisition analyzer 8 in step 30) to obtain the first time domain signal ζ1(t) and the second time domain signal ζ2(t), and obtaining the analysis function R(n) by using the correlation analysis algorithm, as shown in formula (3):

[0049]

[0050] Wherein, R(n) is the cross-correlation analysis function, N is the total number of moments, m represents the moment, and n represents the interval between two signals; ζ1(m) represents the first time domain signal of a speed sensor at moment m; ζ2(m+n) represents the second time domain signal of another speed sensor at moment m+n; when the function R(n) takes the maximum value, the corresponding time interval n=Δt is obtained, and the time difference experienced by the signal to be transmitted to the left and right speed sensors is Δt.

[0051] Step 50) Fault location: Assuming that the propagation speed of the mechanical wave in the rotating shaft (1) is V, and the distance between the speed sensors at both ends of the rotating shaft (1) is S, according to the time difference Δt, the position x where the friction impact occurs is as shown in formula (4):

[0052] x=S / 2-VΔt / 2 Formula (4)

[0053] In formula (3), the position x where the friction impact occurs is the speed sensor on the left side of the shaft as the coordinate zero point.

[0054] The detection device used in the method of this embodiment includes a vibration analyzer and two vibration velocity sensors. The two vibration velocity sensors should be installed at the left and right ends of the fault to be detected respectively; the two vibration velocity sensors are connected to the same vibration analyzer to collect signals.

[0055] 1 is a detection system of the rotating machinery to be detected. The detection system includes a left support bearing seat 2, a right support bearing seat 5, a left vibration velocity sensor 3, a right vibration velocity sensor 6, a computer 7, and a vibration signal analysis and acquisition instrument 8. The left support bearing seat 2 and the right support bearing seat 5 are installed at both ends of the rotating machinery 1, and the left vibration velocity sensor 3 and the right vibration velocity sensor 6 are respectively installed on the holes of the left support bearing seat 2 and the right support bearing seat 5, and are respectively connected to the vibration signal analysis and acquisition instrument 8 through data connection lines, and the vibration signal analysis and acquisition instrument 8 sends data to the computer 7 through the connection line. The computer 7 calculates the location 4 where the friction impact fault occurs.

[0056] The method of the present invention reconstructs the waveform of the signal, which can avoid the interference of the unbalanced force of the rotating machinery itself. Different from using the thermal shock effect of collision and friction to cause the rotor to produce thermal deformation and then diagnose based on the equivalent unbalanced force generated by the deformation, the frequency characteristics of the force impact effect generated by collision and friction are used to reconstruct the signal in step 30). The signals collected by the left vibration velocity sensor 3 and the right vibration velocity sensor 6 contain vibrations at the natural frequency caused by the force impact effect. It is difficult to directly separate the required signal from the complex time domain signal, but after converting it into a frequency domain signal, the collected signal can be screened according to the natural frequency characteristics to separate the required signal.

[0057] The principle of this method is easy to understand, and the installation and operation of the detection device are simple. It is only necessary to install two vibration velocity sensors on the bearing seat and connect the sensor signals to the vibration signal analysis and acquisition instrument connected to the computer. In step 40), the time difference can be obtained by correlation analysis using the two reconstructed signals, and then the fault location can be obtained based on the propagation speed of the mechanical wave in the solid. The calculation is simple and avoids unnecessary and complicated calculation processes.

[0058] The following is a simulation calculation example to illustrate the feasibility of the method of the present invention. Assume that the time domain signals corresponding to the discrete points collected by the vibration velocity sensors at the left and right ends are:

[0059] y1(t)=0.5sin(2π×11t+0.02)+e -t sin(2π×25t)+0.3sin(2π×36t+0.5)

[0060] y2(t)=0.5sin(2π×11t+0.0202)+e -t sin(2π×25t+2×10 -4 )+0.3sin(2π×36t+0.5002)

[0061] The collected signal is reconstructed according to the frequency characteristics of the force impact effect. The time domain signal is subjected to fast Fourier transform to obtain the corresponding frequency domain signal. In this example, the natural frequency of the rotating machinery is known to be 25Hz. Reconstruction screening is performed based on this frequency characteristic, and the range of the reconstruction frequency is set to 20-30Hz, that is, only the frequency domain signal between 20-30Hz is retained, and the amplitudes of the frequency domain signals corresponding to the remaining frequencies (0-20Hz and 30-50Hz in this example) are all set to 0. In this way, only the frequency domain signal at the natural frequency is retained. After that, the processed frequency domain signal is inversely transformed again to obtain the time domain signals ζ1(t) and ζ2(t) reconstructed by the left and right sensors.

[0062] In this example, the total number of time N collected by the vibration signal acquisition analyzer is 1000. Substituting ζ1(t) and ζ2(t) into formula (3), the calculation error at the two end points is large. After eliminating it, the horizontal coordinate value corresponding to the maximum value of R(n) is the time difference: Δt = 4×10 -4 s.

[0063] The distance between the left and right speed sensors is 10m, and the propagation speed of the mechanical wave in the rotating machine 1 is 5000m / s. The coordinates of the fault location are calculated by formula (4): x = 10×0.5-5000×4×10 -4 ×0.5=3m, then the friction impact fault location 4 can be roughly determined to be 3m to the right of the sensor as the coordinate zero point.

[0064] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above specific embodiments. The above specific embodiments and the description in the specification are only for further illustrating the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of the present invention to be protected is defined by the claims and their equivalents.

Claims

1. A method for detecting the position of a rotating machinery collision and impact fault, characterized in that: The following steps are involved: Step 10) installing the detection component; Step 20) When an abnormal friction impact occurs, extract a signal from the rotating machine to obtain a time domain signal; Step 30) reconstructing the time domain signal; Step 40) Analyze the signal to obtain the time difference Δt experienced by the signal being transmitted to the left and right speed sensors; Step 50) Fault location: Assuming that the propagation speed of the mechanical wave in the rotating shaft (1) is V, and the distance between the speed sensors at both ends of the rotating shaft (1) is S, according to the time difference Δt, the position x where the friction impact occurs is as shown in formula (4): x=S / 2-VΔt / 2 Formula (4) In formula (4), the position x where the friction impact occurs is the speed sensor on the left side of the shaft as the coordinate zero point; The step 10) specifically comprises: installing a speed sensor on the bearing seats at the left and right ends of the rotor shaft (1) of the rotating machine to be tested; when installing the speed sensors, the fault position is between the positions of the two speed sensors; The step 20) specifically comprises: when an abnormal friction impact occurs on the rotor shaft (1), the signals measured by the two speed sensors are respectively input into a vibration signal acquisition and analysis instrument (8), and the signal acquisition and analysis instrument (8) is used to record the time domain signal after the friction impact occurs; The step 30) specifically includes: the time domain signal includes multiple single waveforms with different amplitudes, phases, and frequencies, as shown in formula (1): Where y(t) is the vibration displacement of the time domain signal, A1 is the amplitude of the first time domain signal, A2 is the amplitude of the second time domain signal, and A i is the amplitude of the i-th time domain signal, ω1 is the frequency of the first time domain signal, ω2 is the frequency of the second time domain signal, ω i is the frequency of the i-th time domain signal, is the phase angle of the first time domain signal, is the phase angle of the second time domain signal, is the phase angle of the i-th time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, The phase of the sinusoidal signal is damped; According to the frequency characteristics of the force impact of the friction, the signals collected by the two velocity sensors are screened and reconstructed respectively. The reconstruction process is as follows: first, the time domain signal is converted into the corresponding frequency domain signal H(f); the upper limit ωmax and lower limit ωmin of the frequency reconstruction are set in the frequency domain signal H(f) to satisfy the condition ωmin<ω0<ωmax, and the other signal frequencies ω1, ω2, ω i are all outside the interval [ωmin, ωmax]; after conversion, the frequencies ω1, ω2, ω i The signal strength of the corresponding signal is represented by A1, A2, A i Convert to B1, B2, B in the frequency domain representation H(f) i , and then the signal strengths B1, B2, B i It is set to 0, and the frequency domain signal strength corresponding to the interval [ωmin, ωmax] where ω0 is located after the conversion is not changed, so that H(f) can be reconstructed into a new frequency domain signal h(f), in which only the signal corresponding to ω0 exists. Finally, h(f) is inversely transformed into the reconstructed time domain signal. After waveform reconstruction, only the time domain waveform caused by abnormal friction impact is retained, as shown in formula (2): ζ(t)=ζ0e -βt sin(2π×ω0t) formula(2) Where ζ(t) is the vibration displacement of the reconstructed time domain signal, t is the time variable, e is the natural exponent, ζ0e -βt is the amplitude of the damped sinusoidal signal, ω0 is the frequency of the damped sinusoidal signal, The phase of the sinusoidal signal is damped; The step 40) comprises: reconstructing the two speed sensor signals y1(t) and y2(t) collected by the signal acquisition analyzer (8) in step 30) to obtain the first time domain signal ζ1(t) and the second time domain signal ζ2(t), and obtaining the analysis function R(n) by using the correlation analysis algorithm, as shown in formula (3): Wherein, R(n) is the cross-correlation analysis function, N is the total number of moments, m represents the moment, and n represents the interval between two signals; ζ1(m) represents the first time domain signal of a speed sensor at moment m; ζ2(m+n) represents the second time domain signal of another speed sensor at moment m+n; when the function R(n) takes the maximum value, the corresponding time interval n=Δt is obtained, and the time difference experienced by the signal to be transmitted to the left and right speed sensors is Δt.

Images