The vibration diagnosis method for crack & breakdown of roller

Vibration diagnosis method for deterioration and failure of roller rotating parts

The present invention relates to a vibration diagnosis method for the deterioration and failure of a roller rotating mechanism, and particularly relates to a vibration diagnosis method for the deterioration and failure of a feed roller. It mainly uses the characteristics of the vibration signal of the feed roller to determine whether there are abnormal phenomena such as transmission tooth wear, bearing seat looseness, and bearing wear.

According to the key, the feed roll is an important equipment in the rolling process. Its function is to correctly lead the material to be rolled into the work rolls of the rolling mill and maintain a stable feed during the rolling process. If the feed roller does not work properly, it will affect the rolling quality. If there is a sudden failure without warning, the entire production line will be stopped, which will not only cause considerable shutdown losses, but also disrupt the entire maintenance schedule. Form a waste of manpower and material resources.

The feed rollers are located on the left and right sides of the rolling mill, and are divided into an inlet end and an outlet end according to the traveling direction of the steel plate. Figure 1 shows the position of the feed roll in the rolling mill, where 1 is the rolling stand, 2 is the transfer roll table, 3 is the feed roll, 4 is the upper back roll, 5 is the upper work roll, and 6 is The lower work roll, 7 is the lower back roll. A set of feed rollers consists of two rollers, and each roller is supported by two bearings. The two rollers are driven by separate motors and are driven by the way that the drive shaft is directly embedded in the neck of the roller. Photo 1 shows the structure of the four internal transmission teeth of the roller neck.

In the past few years, the feed roll of the rolling mill of the steel plate factory has experienced many sudden failures, causing major property losses. Several bearing condition detection instruments have been tried on site in order to provide early warning before equipment failures and avoid unplanned downtime, but the results have always been ineffective. The reason is that the conventional bearing diagnosis technology uses the characteristic frequency method to detect defects (such as cracks, spalling, etc.) in the bearing components. The diagnosis rules are mostly derived from small test devices in the laboratory and are mostly generated manually. Defects to simulate. Since the abnormal types and conditions of actual bearings are much more complicated than simple artificial defects, there are many shortcomings in applying the vibration characteristics of defects obtained in the laboratory to the diagnosis of actual bearings.

In addition, some instruments on the market for detecting bearing conditions use the principle of ultra-high frequency vibration signals. However, due to the relatively low speed of the feed roller bearing and the use of high-viscosity lubricating grease, the ultra-high frequency signal generated is very weak. It is not easy to be received by the sensor, and the vibration characteristics of some abnormal bearing conditions are manifested in the lower frequency range, but the UHF signal is not easy to detect.

In order to break through this difficulty and bottleneck, the inventor of the present invention set out to develop the feed roller monitoring and diagnosis technology in the steel plate workshop. First, a set of permanently fixed on-line vibration sensors is installed on the feed roller, and then the bearing vibration of the feed roller is measured regularly, and the equipment condition and the characteristics of the vibration signal are constantly compared during the measurement process in order to find the second place. The relationship between the people. After long-term measurement, we finally succeeded in finding out the vibration characteristics of various deterioration conditions of the feed roller (including transmission tooth wear, bearing seat looseness, bearing wear, etc.), and established a vibration diagnosis method for the feed roller condition.

The present invention provides great help for early warning of feed roller failures, not only can extend the service life of the feed roller, reduce maintenance costs, but also can avoid major losses caused by sudden equipment failures. The present invention can be widely applied to other similar equipment, and improve the technical level in the field of equipment monitoring and diagnosis.

First, the measurement device and measurement method are described as follows: The sensor for measuring vibration adopts an accelerometer. The signal generated by it is processed by a charge amplifier, a low-pass filter, and a high-pass filter, and then processed by a signal recorder and a personal computer. Analysis and storage. The signal sampling rate is 5kHz, the cut-off frequency of the low-pass filter is set to 2kHz, and the upper limit of the high-pass filter's frequency is set to 10Hz. After the signal is sampled, the root mean square value and peak value of its amplitude are calculated, and the frequency spectrum is obtained by fast FFT conversion, and the envelope spectrum is found by using the envelope analysis method.

The measurement of vibration is carried out under non-rolling delay and idling of the roll. The reason is: the vibration produced by the steel plate hitting the roll during rolling is much greater than the vibration produced by the rolling of the bearing itself. Therefore, the vibration signal measured during rolling is not helpful for the judgment of the feed roll condition, but only non-rolling The vibration generated by the rotation of the delayed roller itself can reflect the condition of the feed roller.

In order to make the purpose, features, and advantages of the present invention more obvious and understandable, a preferred embodiment and accompanying drawings are described in detail as follows:

Figure 1 shows the position of the feed roller in the rolling mill; Figure 2 shows the vibration signal waveform of the bearing with shock pulse; Figure 3A is a schematic diagram of the gear meshing situation at the beginning of use; Figure 3B is after wear Schematic diagram of the gear meshing situation with increased clearance and eccentricity; Figure 4 shows the relationship between shock wave strength and lubrication status; Figure 5A shows the shock pulse signal; Figure 5B is directly based on Figure 5A to quickly propagate the leaf The converted spectrum; Fig. 5C is the envelope obtained by absolute value and low-pass filtering of the signal of Fig. 5A; Fig. 5D is the envelope spectrum obtained by fast Fourier transformation based on Fig. 5C; Fig. 6A shows Early bearing vibration signal; Fig. 6B is the frequency spectrum of Fig. 6A; Fig. 6C shows the mid-term bearing vibration signal; Fig. 6D is the frequency spectrum of Fig. 6C; Fig. 6E shows the late bearing vibration signal; Fig. 6F is Fig. 6E Figures 7A and 7B show the vibration frequency spectrum of two different bearings in the early stage; Figures 7C and 7D show the vibration frequency spectrum of the bearings in Figures 7A and 7B in the mid-term; Figure 7E and 7F show the bearings in Figures 7A and 7B Vibration spectrum in the later period; Figure 8A is the vibration spectrum when the bearing housing is fixed; Figure 8B is the vibration spectrum when the bearing housing is slightly loosened; Figure 8C is the vibration spectrum when the bearing housing is severely loosened; Figure 9 is the present invention The flow chart for diagnosing the condition of the feed roller; photo 1 shows the four transmission internal teeth of the feed roller neck; photo 2 shows the wear of the feed roller transmission internal teeth; photo 3 shows the wear of the first bearing; photo Four shows the wear condition of another bearing after use.

Three important conclusions can be drawn from the vibration detection results of the feed roll at the exit end of the steel mill: (1) The vibration characteristic of the roller transmission tooth wear is periodic shock pulse waves (2) The vibration characteristic of the bearing wear is vibration. Increase, but the main vibration frequency remains unchanged (3) The vibration characteristic of a loose bearing seat is that the vibration increases and the main vibration frequency decreases. Here are respectively as follows:

(1) The vibration characteristic of the wear of the roller transmission teeth is the appearance of periodic shock pulse waves: Figure 2 shows the vibration signal waveform with shock pulse waves measured at the end of the use of the feed roller. The ones marked with * in the figure are the main shock waves, which should be generated by strong impact; while those without * are the secondary shock waves generated by the main shock waves, whose amplitude is small and attenuates gradually, should be impact The second impact caused by the subsequent rebound.

The main shock wave recurs at a fixed frequency; as shown in Figure 2, the repetition frequency is 7.6 Hz in the preferred embodiment of the present invention. The intensity of the shock wave is also modulated by a frequency that is a fraction of the repetition frequency. In this example, the frequency conversion rate is 1/4 of 7.6 Hz, or 1.9 Hz; this is equivalent to a shock wave that repeats in groups of four, as shown in the figure. As shown by the dotted line, the frequency of the dotted line waveform is exactly the rotation frequency of the roller. Refer to photo 1, the number of transmission teeth is also exactly four, which shows that the external and internal transmission teeth of the shock pulse wave system are generated by impacting each other when they are meshing. Each time the drive shaft rotates, the inner and outer teeth mesh four times, so there is a repetitive pattern of shock waves in groups of four.

When the internal and external teeth are engaged, the impact is caused by the wear of the transmission teeth. Take the schematic diagram of the meshing situation of the transmission teeth in Figures 3A and 3B for illustration: the number 8 represents the neck of the roller, 9 is the transmission shaft, 10 is the external tooth, 11 is the internal tooth, and 12 is the gap. Fig. 3A shows a situation where the degree of meshing is good at the beginning of use; at this time, the torsion force of the transmission shaft 9 is evenly transmitted to the roller 8 via the four external teeth 10. When the transmission teeth begin to wear, the gap 12 between the inner and outer teeth gradually increases. At the same time, because the transmission shaft is not supported by another bearing at the rolling mill end, the gap becomes larger, causing the transmission shaft to sag slightly and form an eccentricity, as shown in Figure 3B. Indicates the situation. This eccentricity causes the uneven distribution of the gap between the inner and outer teeth, so that the torsion force of the drive shaft can only be transmitted by a pair of inner and outer teeth. On the entire feed roller.

After the feed roller was damaged and disassembled, it was found that the transmission teeth on the inner side of the journal were severely worn. Photo 2 shows the situation after the roller is damaged and disassembled. It shows that the transmission internal teeth on the right side have been worn out, so that the power of the motor cannot be transmitted to the roller, and the transmission internal teeth on the left have significant wear (compare Photo 2 and Photo 1). ). The inference that the transmission tooth wear produces shock pulse waves can therefore be fully confirmed.

As for the intensity of the shock wave, it is closely related to the lubrication condition: the worse the lubrication condition, the higher the intensity of the shock wave, and the better the lubrication condition, the lower its intensity. This characteristic can be fully proved by Figure 4: It can be clearly seen from the figure that the shock pulse measured when the rolling is stopped for a period of time (for example, fixed repair) or when the oil is supplied manually before the measurement The intensity is low. The reason is that the lubrication system is not turned off when the rolling is stopped, and it continues to pump grease into the bearings and transmission teeth, and the newly injected grease has the effect of pushing out the old grease and debris. Therefore, the lubrication conditions are better under this condition; especially when the oil is supplied manually, a large amount of grease is injected at a time, and most of the old grease is replaced at one time, so the lubrication effect obtained is the best; as can be seen from the figure , In this case, the vibration is greatly reduced. Therefore, a critical value of shock pulse intensity can be determined; when it exceeds this value, it means that the lubrication is not good; when it is lower than this value, it means that the lubrication is good. As shown in Figure 4, this critical value can be set as 2G in the preferred embodiment of the present invention (G is the acceleration due to gravity, approximately 9.81 meters per second²)。

The repetition frequency of the shock pulse wave cannot be obtained by the general spectrum conversion method, so the analysis method of the envelope spectrum (Envelop spectrum) must be adopted; the signal processing method is as follows: the signal is first half-wave rectified, and then low-pass filtered The high-frequency part is filtered out, leaving the contour of the shock wave (the so-called envelope), and then spectrum conversion is performed on the envelope to obtain the envelope spectrum. From the position of the highest peak in the spectrum, the repetition frequency of the shock pulse can be obtained. Take the example in Figure 5 for illustration: Figure 5A is an impulse pulse signal; if it is directly converted to the spectrum, the spectrum as shown in Figure 5B will be obtained. This frequency spectrum is all high-frequency components, which represent the natural vibration frequency of the bearing caused by the impact of the shock wave, and the repetition frequency of the shock wave itself is quite low-frequency, which cannot be seen in this frequency spectrum. Figure 5C is the envelope obtained by the above-mentioned absolute value and low-pass filtering; it can be clearly seen that what it represents is the contour of the shock wave. After performing frequency spectrum conversion on the waveform of this profile, the envelope spectrum of the 5D image can be obtained. The highest peak in Figure 5D (in this example, 7.6 Hz) is the repetition frequency of the shock wave in Figure 5A, and its basic repetition frequency (in this example, 1.9 Hz, which is 1/4 of 7.6 Hz) And its high harmonics are also clearly visible.

The conventional envelope analysis is carried out by using a commercially available envelope analyzer, which is a hardware device, that is, the signal passes through a series of analog circuits to obtain its envelope. The envelope analysis in the present invention is to digitize the signal first, and then use numerical methods to perform the above-mentioned half-wave rectification (that is, taking the absolute value), low-pass filtering, and frequency spectrum conversion. The envelope curve analysis of this digital method can achieve the same effect as the analog circuit, and it can be achieved with a normal personal computer, so it can save the high cost of instrument purchase.

(2) The vibration characteristic of bearing wear is that the vibration increases, but the main vibration frequency remains unchanged: with the increase of time (wear amount), the bearing vibration gradually increases, but the main frequency components of the vibration do not change much. Figure 6 shows the comparison of the vibration signal and frequency spectrum of the same bearing in different periods. Figures 6A, 6C, and 6E represent the early, mid and late vibration signal waveforms, respectively, and Figures 6B, 6D, and 6F are the frequency spectrums of Figures 6A, 6C, and 6E, respectively. It can be clearly seen that the amplitude of the vibration signal and the area under the spectrum curve increase with the use of time, but the distribution of frequency components remains roughly unchanged.

Comparing different bearings, it can also be found that the greater the wear, the greater the vibration. The two bearings in the preferred embodiment of the present invention are compared. Figure 7 shows the vibration spectrum of the two bearings in the early, middle, and late stages respectively. Figures 7A, 7C, and 7E represent bearings A, and Figures 7B, 7D, and 7F represent bearings B. It can be seen from the figure that at the beginning, the amplitudes of the two bearings are similar (compare Figures 7A and 7B), but the vibration of the A bearing rises rapidly, which obviously exceeds that of the B bearing in the mid-term (compare Figures 7C and 7D) In the late stage, it is much higher than the B bearing (compare the 7E and 7F drawings). Photo 3 and Photo 4 respectively show the internal conditions observed after disassembly of bearing A and bearing B. Photo 3 shows a large amount of wear in the A bearing. A half-circle of the retainer was worn out and the gap was too large, so that the rollers could not be held in it. Therefore, when the bearing was disassembled, the rollers were scattered. Photo 4 shows that the condition of the B bearing is very good, and its roller surface and ring surface are still smooth and smooth, indicating that the amount of wear is not large. The above comparison confirms that the amount of vibration can indeed effectively reflect the degree of wear of the bearing.

(3) The vibration characteristic of bearing housing looseness is that the vibration increases and the main vibration frequency decreases: when the bearing housing gradually loosens, in addition to the increase in the vibration value, the vibration frequency also continues to decrease. Figure 8 shows the vibration spectrum of the same bearing under different degrees of looseness. Figure 8A is the vibration spectrum when the bearing housing is fixed, Figure 8B is the vibration spectrum when the bearing housing is slightly loosened, and Figure 8C is the vibration spectrum when the bearing housing is severely loosened. The vibration spectrum. It can be found from the figure that the more stable the bearing seat, the higher its vibration frequency, and the looser the bearing seat, the lower its vibration frequency. Therefore, a quantitative index that can represent the degree of looseness of the bearing seat is developed in the present invention, that is, the center frequency of the vibration spectrum, which is defined as:<maths><img file="TW286359B_D0001.tif" /></maths>Where fc = the center frequency of the vibration spectrum, which is equivalent to the center of gravity of the area under the spectrum line

f=The frequency of each spectrum line in the vibration signal spectrum

s(f)=The height of each spectrum line in the vibration signal spectrum

The center frequency of the vibration spectrum is an important indicator for judging whether the bearing seat is loose. The higher the value, the better the degree of fixing of the bearing seat; the lower the value, the looser the bearing seat. The accuracy and reliability of this judgment index have been fully verified after repeated verifications on site.

Based on the above analysis, the present invention has developed a complete set of feed roller condition diagnosis method, and the process is shown in Figure 9. The diagnosis process is as follows: Firstly, determine whether the amplitude exceeds the set critical value. If the vibration is lower than this value, it is judged that the feed roller is normal; if it is higher than this value, it is judged that the feed roller is abnormal. This threshold is set at 150mG in the preferred embodiment of the present invention. Secondly, using the waveform characteristics of the vibration signal, the abnormal conditions are divided into two categories. One is with shock pulses, such as transmission tooth wear and bearing component damage, and the other is without shock pulses, such as loose bearing housings and bearings. Internal wear. The difference in the waveform characteristics of the two can be expressed by a quantified index, namely the crest factor, which is defined as the comparison between the maximum peak value of the signal and its root mean square value. According to the statistical analysis of hundreds of sets of measurement data obtained in the research and development process of this case, the following results can be obtained: the wave height of the wave with shock pulse wave is above 5, and the wave height of the wave without shock pulse wave The rates are all between 3 and 4; therefore, the constant wave height rate of 4.5 can be used to distinguish the boundary between the two.

The next classification is based on the frequency characteristics of the signal; the waveform with shock wave uses envelope spectrum analysis to find out the repetition frequency of the shock wave, and judge whether this frequency is the characteristic frequency of transmission tooth wear or bearing component damage, and then judge the abnormality separately It is due to abrasion of transmission gears or damage to bearing components. For waveforms without shock waves, general spectrum analysis is used to find the center frequency, and based on this frequency system is less than the lower limit of the frequency or greater than the upper limit of the frequency, or between the two, it is judged that the cause of the abnormality is the looseness of the bearing seat, Internal wear of the bearing or a combination of both. The lower limit and upper limit of the frequency are respectively set at 500 Hz and 600 Hz in the preferred embodiment of the present invention.

The invention is the result of long-term vibration measurement and analysis of the feed roller in actual operation, and it can successfully diagnose the abnormal phenomena of the feed roller including the wear of the transmission gear, the looseness of the bearing seat, and the wear of the bearing. The correctness of this diagnostic method has been repeatedly verified and has been fully confirmed. After the equipment was damaged and disassembled, it was also found that the actual damage situation was exactly the same as that predicted by the diagnosis method. The newly installed feed roller was found to follow the same deterioration development process, which confirmed the universal applicability of this diagnostic method.

Although the present invention has been disclosed as above in a preferred embodiment, it is not intended to limit the present invention. Anyone who is familiar with the art can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, The protection scope of the present invention shall be subject to those defined by the attached patent application scope.