This paper focuses on the bearing cage fracture failures occurred in the radial bearings of the distribution shafts of the 16# and 17# rolling mill gearboxes in Masteel Special Steel High-Speed Wire Rod Mill. It analyzes the reasons why the online vibration monitoring sensors failed to capture the abnormal vibration signals in advance during the later two failures after the first failure and the installation of vibration sensors. By comparing the situations before and after the replacement of the bearing cage material, it also explores the effectiveness of the material change. Through detailed analysis of the vibration data and the characteristics of the bearing failures, this paper aims to provide valuable references for similar problems in the industry.
1. Introduction
In the production process of the special steel high-speed wire rod mill, the reliability of the rolling mill gearbox is crucial. In April 2024, the 16# and 17# rolling mill gearboxes in Masteel Special Steel High-Speed Wire Rod Mill had abnormal noises. After inspection, it was found that the bearing cages of the radial bearings of the 16# and 17# distribution shafts were damaged. After the first repair, vibration sensors were installed for monitoring, but in the subsequent two failures, the online vibration monitoring system did not capture the abnormal vibration signals in advance. This situation has brought great challenges to the stable operation of the equipment and the prediction of failures.
2. Equipment Structure and Monitoring Point Layout
2.1 Equipment Composition
The 16# and 17# rolling mills mainly consist of motors, gearboxes, and roll boxes. The motor power is 1500 kW. The gearbox plays a key role in transmitting power and torque, and the bearings in it are important components to ensure the normal operation of the shaft. The structure diagram of the motor and gearbox is shown in Figure 1 (Here, it is assumed that Figure 1 can clearly show the connection and structure of each part).
2.2 Initial Monitoring Point Layout
After the first bearing fracture failure, due to the limitation of the bearing positions of the 16# and 17# distribution shafts, only the axial measuring points of the gearbox input shaft and the measuring points of the 16# and 17# roll box bodies were monitored for vibration. This layout may not be able to accurately capture the vibration signals of the key bearings, resulting in the inability to predict the bearing failures in time.
2.3 Optimized Monitoring Point Layout
After the second failure, the monitoring points were optimized. A total of 7 vibration sensors were installed at the axial drive end of the gearbox input shaft, the axial free end of the input shaft, the horizontal side of the 16# distribution shaft drive side near the bearing rib plate, the axial free end of the 16# distribution shaft, the axial drive side of the 17# distribution shaft, and the vertical and axial directions of the 17# distribution shaft free end near the bearing rib plate. This layout aims to more comprehensively monitor the vibration of the gearbox and improve the accuracy of failure prediction.
3. Vibration Analysis of Gearbox Bearings
3.1 Vibration Status after the First Bearing Replacement
After the gearbox bearing was replaced on April 17, 2024, the vibration acceleration waveform had impact signals with the rotation frequencies of the 16# and 17# distribution shafts as the fundamental frequencies, and the vibration acceleration value was 2.3 m/s². In the frequency spectrum, there were also equally spaced impact signals. This indicates that there may be problems such as excessive clearance in the bearing, although the vibration amplitude was relatively small at this time. The vibration acceleration waveform and frequency spectrum monitoring data of the axial direction of the gearbox input shaft on April 20, 2024 are shown in Figures 4 and 5 (Here, Figures 4 and 5 can visually display the waveform and frequency characteristics).
| Date | Vibration Acceleration Value (m/s²) | Impact Signal Characteristics |
|---|---|---|
| April 17, 2024 | 2.3 | With the rotation frequencies of 16# and 17# distribution shafts as the fundamental frequencies |
| April 20, 2024 | (Details in Figures 4 and 5) | (Shown in the figures) |
3.2 Vibration Status before the Second Bearing Failure
Before the 17# distribution shaft bearing failed on May 26, 2024, the speed and acceleration trends of the axial measuring points of the gearbox input shaft were stable, and the vibration amplitudes were small. The vibration speed amplitude was within 0.6 mm/s, and the acceleration amplitude was within 3 m/s². In the vibration monitoring frequency spectrum, there was an equally spaced impact signal at 28.125 Hz. In the envelope demodulation spectrum of the axial vibration acceleration frequency spectrum of the gearbox input shaft, there were the same impact frequency as the rotation frequency of the 17# distribution shaft (28.125 Hz) and its second, third, and fourth harmonics, but there was no obvious abnormal signal. The relevant data trends are shown in Figures 6-9.
| Date | Vibration Speed Amplitude (mm/s) | Vibration Acceleration Amplitude (m/s²) | Impact Signal Frequency (Hz) |
|---|---|---|---|
| May 26, 2024 | < 0.6 | < 3 | 28.125 |
3.3 Vibration Status after Adding Sensors for the Second Time
After adding 7 vibration sensors on May 28, 2024, according to the vibration data monitored on May 30 and 31, there were obvious impact signals of the 16# and 17# distribution shafts in the gearbox, and the impact signal characteristics were obvious. In the envelope demodulation spectrum, there were the same impact frequency as the rotation frequency of the 17# distribution shaft (28.438 Hz) and its second, third, and fourth harmonics. This shows that the optimized monitoring point layout can capture more vibration information, but it still cannot accurately predict the bearing cage fracture failure. The relevant waveforms and frequency spectra are shown in Figures 10-11.
3.4 Vibration Status before the Third Bearing Failure
On June 27, 2024, when the 17# distribution shaft bearing had a serious failure of the cage fracture again, the vibration speed value of the axial measuring point of the 17# distribution shaft free end was small, and the acceleration value was also small but had a slight increasing trend. The vibration speed value of the axial measuring point of the input shaft free end was small, the acceleration value had an increasing trend and a relatively large amplitude, but the vibration waveform and frequency spectrum had no obvious abnormal signals. The data trends are shown in Figures 12-17.
| Measuring Point | Vibration Speed Trend | Vibration Acceleration Trend |
|---|---|---|
| 17# Distribution Shaft Free End Axial | Small, slight increase | Small, slight increase |
| Input Shaft Free End Axial | Small | Increase, large amplitude |
4. Analysis of the Causes of Bearing Cage Failure
4.1 Characteristics of Bearing Cage Failure
The bearing cage fracture failure is different from the failures of the inner ring, outer ring, and rolling elements of the bearing. The failures of the inner ring, outer ring, and rolling elements usually take a certain time to form, and during this process, the vibration value of the bearing will gradually increase, and there will be obvious characteristic frequencies of the bearing element failures in the vibration waveform and frequency spectrum signals, which are easy to identify. However, the characteristic frequency of the cage failure is the smallest among the failure frequencies of all bearing elements, and it is difficult to identify. In addition, the bearing cage fracture occurs instantaneously. Before the failure, the bearing state is basically normal, and the rolling elements are also operating in an orderly manner. Once the cage breaks, it is already in the later stage of the bearing failure, and although the acceleration amplitude will rise in a short time, it is difficult for the online vibration monitoring system to capture the abnormal vibration signal in advance.
4.2 Influence of Sensor Layout
The initial sensor layout did not cover the key positions of the bearings, resulting in the inability to accurately capture the vibration signals related to the bearing cage. Even after the optimization of the monitoring points, due to the characteristics of the cage failure itself, it was still difficult to predict the failure in advance. This shows that in the design of the monitoring system, it is necessary to fully consider the characteristics of different bearing failures and select the most appropriate sensor layout positions.
4.3 Material Factors
After replacing the bearing cage from steel to copper with stronger plasticity and ductility, the unit operated for four months without any fracture of the bearing cage. This indicates that the material properties of the bearing cage have a significant impact on its service life. The steel material may have insufficient plasticity and ductility, resulting in easy fracture under certain working conditions, while the copper material can better adapt to the stress changes during the operation of the bearing and reduce the probability of fracture.
5. Conclusion and Prospect
5.1 Conclusion
In this paper, through the analysis of the bearing cage fracture failures of the 16# and 17# rolling mill gearboxes in Masteel Special Steel High-Speed Wire Rod Mill, it is found that the difficulty in predicting the bearing cage failure by vibration monitoring is due to the instantaneous nature of the failure and the small characteristic frequency. The sensor layout also has an important impact on the monitoring effect. In addition, the selection of the bearing cage material is crucial to the reliability of the bearing. By replacing the steel cage with a copper cage, the problem of frequent cage fractures has been effectively solved.
5.2 Prospect
In future research, more advanced vibration monitoring technologies and signal processing methods can be explored to improve the ability to capture and analyze weak failure signals. At the same time, further research on the material performance and design of the bearing cage can be carried out to continuously improve the reliability and service life of the bearing. In addition, for the layout of the monitoring points, more simulation and experimental studies can be carried out to find the optimal layout scheme under different equipment structures and working conditions.
