CN101556200B - Vector spectrum based dynamic balance method for flexible rotor - Google Patents

Abstract

The invention discloses a flexible rotor dynamic balance method based on full vector spectrum, through which full vector spectrum analysis software can objectively and truly find rotor unbalance faults on site, and by measuring For the original vibration signal, use the full vector spectrum filtering method to extract the power frequency main vibration vector of the rotor at each vibration measurement section, and use this main vibration vector to replace the rotor power frequency amplitude and phase extracted from the traditional single-channel signal, combined with the influence coefficient method On-site dynamic balancing, the balancing process is less dependent on the operator's experience and knowledge, which can significantly improve the efficiency and accuracy of flexible rotor dynamic balancing. It is a highly integrated computer-aided on-site dynamic balancing method.

Description

Dynamic Balancing Method of Flexible Rotor Based on Full Vector Spectrum

technical field

The patent of the present invention belongs to the technical field of on-site dynamic balancing of rotors, and specifically relates to the field of fault diagnosis and vibration control of rotating machinery. the

Background technique

At present, there are many on-site dynamic balancing methods for flexible rotors, which can be roughly classified into two categories, namely, the mode balance method and the influence coefficient method. Separately balanced, it has the characteristics of high-speed balancing accuracy and less test times. However, the vibration-shape balancing method requires first to obtain the vibration-shape data of each order of the rotor. At the same time, this method requires the operator to have rich experience and knowledge, and it is not easy. The use of computer-aided balancing work is severely limited in practical applications. The influence coefficient method is less dependent on the operator's experience and knowledge, and the entire operation is completed by experiments, which can be computer-assisted and automated, so it is very popular in many occasions. The disadvantage of the influence coefficient method is that it is not sensitive to the high-order mode shapes of the rotor. The calculation of the influence coefficient mainly depends on the accuracy of the vibration amplitude and phase measurement. Impossible to implement on site. Therefore, the measurement and calculation accuracy of various balance parameters, especially the accurate acquisition of vibration amplitude and phase information, is particularly important for the balance quality. However, in actual balance, due to the many links in the generation and transmission of balance errors, which are not easy to control, a small amount of original errors may eventually produce large balance errors after accumulation, resulting in balance failure. the

Contents of the invention

The purpose of the patent of the present invention is to provide a flexible rotor dynamic balance method based on the full vector spectrum, which can objectively and truly find the unbalanced fault of the on-site rotor through the full vector spectrum analysis method. The original vibration signal of each channel is combined with the influence coefficient method for on-site dynamic balance, which significantly improves the efficiency and accuracy of the flexible rotor dynamic balance.

In order to achieve the above object, the technical solution of the present invention is: a method for dynamic balancing of flexible rotors based on full vector spectrum, the steps of the method are as follows:

1) According to the main vibration vector diagram of the full vector spectrum to identify whether there is an unbalanced fault in the rotor, according to the dual-channel eddy current signal collected on the cross-section of the upper rotor of the unit, perform vector spectrum analysis, and directly determine the rotor’s fault from the analysis spectrum. Rotor unbalance fault; the method of discrimination is: find a spectral line with the highest amplitude from the main vibration vector diagram of the vector spectrum. If the standard vibration value exceeds the standard, the possibility of rotor imbalance is extremely high, and then judge from the other end of the rotor. If the situation is similar, it can be judged as a rotor imbalance fault, and on-site rotor dynamic balancing is required;

2) Select at least two weighted planes according to the structure of the rotor; when balancing on site, determine the number of weighted planes and the position of the weighted planes according to the balance effect and actual conditions. the

3) Arrange sensors. The number of sensors is at least twice the number of weighted planes. At least one pair of vibration sensors is arranged on each side of the rotor. Each pair of vibration sensors is placed perpendicular to each other near the outer surface of the rotor shaft to measure , the horizontal and vertical vibration near the right bearing, and a key phase sensor is placed on the outer circumference of one of the weighted planes for the measurement of the key phase signal;

4) Set the weighting radius on each weighting plane, add a test weight, start the machine to the set speed, and measure the corresponding main vibration vector;

5) The counterweights that should be added on each weighted plane can be calculated by using the influence coefficient method for each main vibration vector. the

Further, the method also has a step of balance verification. the

Further, if there is still an imbalance, repeat steps 4), 5) and balance verification steps until the balance reaches the standard. the

In general, the sensors are symmetrically arranged at both ends of the rotor, and the sensors for the dynamic balance of the double-weighted plane rotor are placed perpendicular to each other near the outer surface of the rotating shaft at both ends of the rotor, and are used to measure the horizontal and vertical values near the left and right bearings. vibration. the

The vibration sensor uses a photoelectric sensor or an eddy current sensor to measure the speed of the rotor of the machine and obtain the pulse signal of the vibration phase reference. For the photoelectric sensor, it is required to stick a reflective narrow band on the mark on the shaft, and the rest is a black area, or vice versa; For the eddy current sensor, it is required to open a keyway at the marked line on the shaft, and connect the above sensor signal line to the dynamic balancer. the

For rotors with two weighted planes, the balancing method is as follows:

a) Start the machine, increase the speed, start the dynamic balancer to measure the initial vibration of the two planes, and use the vector filter algorithm to directly calculate the initial power frequency main vibration vector of the two planes;

b) For the two weighted planes, set the weighted radius, add a test weight Q ₁ in one of the planes, Q ₁ is a vector, its mass is q ₁ , relative to the azimuth angle of the reference mark on the rotor, the reverse rotation direction is calculated as Start the engine up to the same speed, and measure the corresponding main vibration vector;

c) Remove Q ₁ and add test weight Q ₂ on another plane to measure its corresponding principal vibration vector;

d) Using the influence coefficient method, the counterweights that should be added on the two planes can be calculated. the

The invention patent has the following functions and characteristics:

1. This method provides vector spectrum analysis software and corresponding map display functions. The vector spectrum has the characteristics of high resolution and fusion of dual-channel signals, and has unique advantages for the identification of rotor unbalance faults on site. the

2. The vibration measurement sensor adopts a single-section double-channel arrangement arranged at 90 degrees, and the sensor is a non-contact eddy current displacement sensor to measure shaft vibration. The signal acquisition adopts the method of synchronous whole cycle, which can be triggered by key phase signal or by photoelectric switch. the

The software filtering method based on the vector spectrum is used to extract the main vibration vector of the power frequency of each vibration measurement plane. The filtering method uses the fast algorithm of the complex FFT to synchronize the characteristics of the entire cycle signal acquisition, which improves the accuracy and accuracy of the power frequency amplitude and phase extraction. Efficiency, better avoid the one-sidedness of the single-channel signal, thereby improving the balance accuracy of the influence coefficient method. the

3. The number of vibration measuring planes can be specified arbitrarily, which fully meets the dynamic balance requirements of most flexible rotor shafting systems. the

4. The balancing process is carried out with computer aids, the operation is intuitive and clear, and less dependent on expert experience. the

Description of drawings

Figure 1 is the layout scheme of on-site measuring points;

Fig. 2 is vector spectrogram;

Fig. 3 is a flowchart of vector spectrum filtering. the

The specific implementation mode of the patent of the present invention is:

(1) Rotor unbalance fault identification based on vector spectrum

Detailed ways

According to the dual-channel eddy current signal collected on a section at one end of the upper rotor of the unit, the vector spectrum analysis is carried out, and it can be directly determined whether the rotor belongs to the rotor unbalance fault from the spectrum. The method of discrimination is as follows: a spectral line with the highest amplitude can be found from the main vibration vector diagram of the vector spectrum. If the vibration value exceeds the standard, the rotor is likely to be unbalanced, and then judge from the other end of the rotor. If the situation is similar, it can basically be judged as an unbalanced fault of the rotor, and on-site rotor dynamic balancing is required. the

Refer to Figure 2 for the method of identifying rotor unbalance faults using vector spectrum diagrams. The figure shows a set of vibration data measured by a pair of eddy current sensors x and y perpendicular to each other from a CO ₂ compressor unit site close to the booster box section, and the Spectrum diagram and main vibration vector diagram are made separately in x and y directions. If it can be seen from the amplitude diagram in the x direction alone that the frequency multiplier of the unit is the largest, it is most likely that the rotor is misaligned. Unbalanced faults, that is, there is a contradiction in judging the faults of the unit with the x and y directions alone, which is a problem caused by the one-sidedness of the single-channel signal. The main vibration vector diagram is an amplitude diagram obtained by combining signals in the x and y directions, which is objectively authentic and unique. From the main vibration vector diagram of this set of data, it can be seen that the 1-octave frequency amplitude of the unit is obvious , and the 2 multiplier is small, so it is certain that the rotor imbalance fault has occurred in the unit.

(2) Determination of the weighted plane

During on-site balancing, the number of weighted planes and the weighted positions are generally determined according to the balance effect and actual conditions. the

(3) Arrangement of sensors

The arrangement of sensors can be carried out according to the characteristics of the rotor, and the number of sensors is at least twice the number of weighted planes. The sensors are often arranged symmetrically at both ends of the rotor. The sensor layout of the dual-plane rotor dynamic balance is shown in Figure 1. A total of 5 eddy current sensors are used. The key phase sensor KP is used to measure the key phase signal, 1X, 1Y , 2X, 2Y are used to measure the horizontal and vertical vibration near the left and right bearings respectively. the

(4) Dynamic balancing process

The following uses two planes as an example to illustrate the process of using the on-site dynamic balancer based on the full vector spectrum to achieve balance. The balance of flexible rotors generally requires multi-plane balancing at multiple speeds from low to high. Here, a certain speed n (rev/min) as an example to explain the balance process (the balance process at other speeds is similar):

a) Select two vibration measuring planes 1 and 2 on the machine, and install two eddy current sensors on each plane at 90° to each other, and use photoelectric sensors or eddy current sensors to measure the speed of the machine and obtain Pulse signal for vibration phase reference. If it is a photoelectric sensor, it is required to stick a reflective narrow strip on the mark on the shaft, and the rest is a black area, or vice versa. If an eddy current sensor is used, it is required to open a keyway a few millimeters deep at the marked line on the shaft. Connect the above 5 sensor signal lines to the dynamic balancer. the

b) Start the machine, increase the speed to n rpm, start the balancer to measure the initial vibrations in the x and y directions of the two planes, and use the vector filter algorithm to directly calculate the initial power frequency main of plane 1 and plane 2 Vibration vectors V ₁₀ and V ₂₀ ; the principle of calculation of power frequency main vibration vectors is shown in Figure 3. Take the calculation of the main vibration vector V ₁₀ of plane 1 as an example:

Suppose the original vibration signals in the X and Y directions measured on plane 1 are respectively {xi _} and { _yi } (i=0, 1, 2, ..., N-1), N is the number of signal collection points, generally The integer power of 2, such as 512, 1024, 2048, etc., R is the number of sampling points per cycle, generally 32, 64, etc. Then construct complex signal z _i =xi+jy _i (i=0,1,2,...,N _- 1), and adopt formula (1) to carry out complex FFT transformation to the constructed signal:

$Z_{\mathrm{no}}=\underset{k=0}{\overset{N-1}{\mathrm{\Σ}}}{z}_k{e}^{-j2\mathrm{\πk}/N}$ (n=0, 1, 2,...N-1) (1)

Then, calculate the serial number Ind of the power frequency main vibration vector spectrum:

Ind＝N/R-1 (2)

The power frequency main vibration vector (including amplitude and azimuth angle) can be calculated by formulas (3) and (4):

$R_{\mathrm{ai}}=(\sqrt{Z_{\mathrm{II}}^2+Z_{\mathrm{Ri}}^2}+\sqrt{Z_{\mathrm{Ri}}^2+Z_{I(N-i)}^2})/2N$ (i=Ind) (3)

${\mathrm{\α}}_i=\arctan \left(\frac{Z_{\mathrm{Ri}}{Z}_{\mathrm{II}}+Z_{I(N-i)}{Z}_{R(N-i)}}{Z_{\mathrm{Ri}}{Z}_{R(N-i)}-Z_{\mathrm{II}}{Z}_{I(N-i)}}\right)$ (i=Ind) (4)

Then the initial power frequency principal vibration vector V ₁₀ of plane I =R _aInd ∠α _Ind

)，同样，启机升至同一转速，并测得相应的主振矢为V₁₁和V₂₁；  c) Select two weighted planes 1 and 2 according to the structure of the rotor, the weighted radii are r ₁ and r ₂ respectively, first add a test weight Q ₁ in plane 1, (Q ₁ is a vector, its mass is q ₁ , Relative to the azimuth angle of the reference mark on the rotor, the angle of the reverse direction of rotation is

), similarly, start the machine up to the same speed, and measure the corresponding main vibration vectors as V ₁₁ and V ₂₁ ;

At this time, the influence coefficient (the effect vector of the unit test weight with the azimuth angle being zero degrees) α ₁ and β ₁ can be calculated:

${\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}}{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 20</span>}}}{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

d) Remove Q ₁ and add test weight Q ₂ on plane 2 to measure V ₁₂ and V ₂₂ ;

Similarly, α ₂ and β ₂ can be calculated:

${\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}}{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 20</span>}}}{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

e) Using the influence coefficient method, it can be calculated that the counterweights that should be added on plane 1 and plane 2 are P ₁ (including the size and direction of the counterweight, the definition of the positive direction is the same as that of adding test weights Q ₁ and Q ₂ ) and _P2 . The way to calculate _P1 and _P2 is to solve the vector equation system (9):

$\left\{\begin{array}{c}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="P"\; filenames="H2009101432247E00021.png"\; state="\{\{state\}\}">P</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\\ {\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="P"\; filenames="H2009101432247E00021.png"\; state="\{\{state\}\}">P</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; 20</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

f) Add counterweight according to the calculation result of step (e), start the machine to the same speed, measure the remaining vibration vectors V′ ₁ and V′ ₂ , such as V′ ₁ and V′ ₂ relative to the initial vibration vectors V ₁₀ and V ₂₀ The amplitudes have been greatly reduced, indicating that the balance effect is better.

(5) Balance verification and conclusion

In order to verify the balance effect of the full vector dynamic balance, a lot of verification work has been carried out on the rotor test bench, and the experimental process is carried out according to the above implementation steps. The experiment compared the full vector dynamic balancing method with the traditional dynamic balancing method based on unidirectional sensor signals, and the experimental data are listed in Table 1. the

Table 1 Dual-plane dynamic balance experiment data (balance speed: 2400 rpm)

Table 2 Calculation and Balance Verification

It can be seen from Table 2 that the counterweight calculated by the full vector dynamic balance method is slightly different from the counterweight calculated by the X or y direction sensor signal method alone. In order to verify the accuracy of various dynamic balance methods, respectively Add the counterweight according to the three counterweight schemes, and measure the vibration in the x and y directions of the two planes after adding the counterweight. Through the analysis of the vibration data after the counterweight is added, it can be obtained that the effect of the above two balancing methods (using X and Y direction signals alone) is basically the same, and the final vibration is too large. In the case of high balance precision requirements, it is often necessary to further balance. However, the vibration measured by the full vector dynamic balance method is relatively small in both x and y directions, and its balance effect is obviously better than the traditional balance method based on unidirectional signals. the

Through comparative analysis, it can be seen that compared with the traditional method, the full vector dynamic balance method has higher precision, can significantly improve the balance effect and reduce the number of balance times. the

Claims (6)

1. A flexible rotor dynamic balancing method based on full vector spectrum, characterized in that the steps of the method are as follows: 1) According to the main vibration vector diagram of the vector spectrum to identify whether there is an unbalanced fault in the rotor, according to the dual-channel eddy current signal collected on the cross-section of the upper rotor of the unit, perform vector spectrum analysis, and directly determine the rotor of the rotor from the analysis spectrum Unbalanced fault; the method of discrimination is as follows: a spectral line with the highest amplitude can be found from the main vibration vector diagram of the vector spectrum. If the spectral line is 1 times the frequency of the rotor and is obviously higher than other spectral lines, refer to the international vibration standard If the vibration value exceeds the standard, the possibility of rotor imbalance is very high, and then judge from the other end of the rotor. If the situation is similar, it can be judged as a rotor imbalance fault, and on-site rotor dynamic balancing is required; 2) Select at least two weighted planes according to the structure of the rotor; 3) Arrange sensors. The number of sensors is at least twice the number of weighted planes. At least one pair of vibration sensors is arranged on each side of the rotor. Each pair of vibration sensors is placed perpendicular to each other near the outer surface of the rotor shaft for measuring , the horizontal and vertical vibrations near the right bearing, and a key phase sensor is placed on the outer circumference of one of the weighted planes to measure the key phase signal; 4) Set up a weighting radius on each weighting plane, add a test weight, start the machine up to the set speed, and measure the corresponding main vibration vector; 5) The counterweights that should be added on each weighted plane can be calculated by using the influence coefficient method for each main vibration vector. 2. The flexible rotor dynamic balancing method based on full vector spectrum according to claim 1, characterized in that, the method also has a step of balance verification. 3. The dynamic balancing method for a flexible rotor based on full vector spectrum according to claim 2, wherein if there is still an unbalance, repeat steps 4), 5) and balance verification steps until the balance reaches the standard. 4. The flexible rotor dynamic balancing method based on full vector spectrum according to claim 1, 2 or 3, characterized in that the vibration sensors are symmetrically arranged at both ends of the rotor, and the vibration sensors of the double-weighted plane rotor dynamic balance are perpendicular to each other Placed near the outer surface of the rotating shaft at both ends of the rotor, it is used to measure the horizontal and vertical vibrations near the left and right bearings. 5. The flexible rotor dynamic balancing method based on full vector spectrum according to claim 4, wherein the key phase sensor adopts a photoelectric sensor or an eddy current sensor to measure the rotating speed of the machine rotor and obtain the pulse signal of the vibration phase reference, For the photoelectric sensor, it is required to stick a reflective narrow strip on the mark on the shaft, and the rest is a black area, or vice versa; for the eddy current sensor, it is required to open a keyway at the marked line on the shaft, and connect the above sensor signal line to the dynamic balancer . 6. The dynamic balancing method for flexible rotors based on full vector spectrum according to claim 5, characterized in that, for a rotor with two weighted planes, the balancing method is as follows: a) Start the machine, increase the speed, start the dynamic balancer to measure the initial vibration of the two planes, and use the vector filter algorithm to directly calculate the initial power frequency main vibration vector of the two planes; b) For the two weighted planes, set the weighted radius, and add a trial weight to one of the planes

,

is a vector whose mass is

, relative to the azimuth angle of the reference mark on the rotor, the angle of the reverse rotation direction is

, start the machine up to the same speed, and measure the corresponding main vibration vector; c) take away

, add a test weight on another plane

, its corresponding principal vibration vector can be measured; d) Using the influence coefficient method, the counterweights that should be added on the two planes can be calculated.

Images