Abstract
In order to enhance the rationality of gear hobbing process parameter selection, this paper delves into the optimization of process parameters based on vibration signals. By constructing a vibration monitoring system utilizing acceleration sensors, conducting numerous cutting experiments, and analyzing the experimental data, we aim to identify issues in traditional cutting processes and propose optimized process parameters. This study is crucial for improving machining accuracy, reducing tool wear, extending machine tool life, and lowering production costs for enterprises.
1. Introduction
Gears serve as the fundamental transmission components in automotive, engineering machinery, and other drive systems. They are extensively utilized in various machinery due to their constant power output, high load capacity, and efficient transmission. Gear hobbing, characterized by its high efficiency and adaptability, stands out as one of the most widely applied gear manufacturing techniques. However, the majority of gear manufacturing enterprises in China rely on the experience of technicians or reference manuals for selecting process parameters. This approach may overlook factors such as vibration during the machining process, which can lead to decreased machining accuracy, shortened tool life, and reduced machine tool life. Therefore, the rational and scientific selection of process parameters is of great significance in gear hobbing. This paper collects vibration data from different process parameters during gear hobbing and optimizes the process parameters based on these vibration signals.
2. Construction of the Monitoring System
2.1 Introduction to the Gear Hobbing Machine
The gear hobbing process employed in this study utilizes the QinChuan YK3126 gear hobbing machine. This machine boasts seven axes with four simultaneous movements and includes seven CNC axes: hob feed axis X, tool slide axis Y, hob traverse axis Z, tool rotation axis A, tool spindle S (B), workpiece rotation axis C, and manipulator rotation axis C2. Controlled by a numerical control system, the machine can accomplish gear and worm rolling and achieve various tooth profile cutting methods, such as straight tooth cutting, helical tooth cutting, crowned tooth cutting, conical tooth cutting, radial worm rolling, and tangential worm rolling. It can perform both forward and reverse hobbing for gears of different modules and pressure angles.
2.2 Selection and Installation of Vibration Sensors
Due to the limited space available for installing vibration sensors on the machine tool, three-axis Dytran 3143D1 acceleration sensors were selected. These sensors have a measuring range of ±50g, an output sensitivity of 100mV/g, and a frequency response of 10kHz. Being insulation-type sensors, they can be installed directly.
2.3 Overall Architecture of the Monitoring System
Vibration data acquisition was performed using vibration acquisition cards from Huazhong University of Science and Technology Xiangyang Polytechnic Institute. The vibration sensors were connected to a switch through the vibration acquisition module, and a laptop was connected to the switch to run the acquisition software for data collection. The overall architecture of the monitoring system.
3. Cutting Experiments
3.1 Experimental Design
The process parameter optimized in this experiment is the spindle speed, and thus, the relationship between spindle speed and vibration was investigated. The gear hobbing process was divided into two cuts: the first cut to remove material and the second cut to improve machining accuracy. The specific process parameters are shown in Table 1. During the experiments, the axial feed rate and radial cutting depth remained constant, while the tool speed was gradually increased.
3.2 Optimization of Process Parameters Based on Experimental Data
Firstly, the data from the X, Y, and Z directions of part number 1 were analyzed as a whole. It can be observed that the overall trends in the three directions are consistent, with the vibrations in the X and Z directions being larger than those in the Y direction. This indicates that the X and Z directions are vibration-sensitive, as they are the directions of greater force application, while the Y direction, being axial, experiences less force. Next, focusing on the X direction and combining it with the machining process, the machining stages were divided.
To ensure the completeness of the collected machining process, the acquisition software was started before the program was initiated. The vibration was initially small and very stable, indicating that the first stage was when the program was not yet started. In the second stage, the vibration suddenly increased, indicating that the program had started and the spindle was rotating, but had not yet contacted the workpiece. In the third stage, the vibration suddenly increased again as the tool contacted the workpiece. During the latter half of the third stage, as the cutting depth gradually decreased, the vibration also decreased, with a sudden drop at the moment of detachment. The vibration in the fourth stage was similar to that in the second stage, representing the interval between the two cuts. The vibration trend in the fifth stage was similar to that in the third stage, representing the second cutting process. The vibration in the sixth stage was similar to that in the second stage, representing the process of the hob disengaging from the workpiece and retracting.
The spindle speeds for the first cut of parts 1, 2, 3, 4, and 5 gradually increased from 440r/min to 800r/min, and for the second cut, from 570r/min to 850r/min. The axial feed rates for the first and second cuts were 42.18mm/min and 50.62mm/min, respectively, while the radial cutting depths were 6.45mm and 0.3mm, respectively. Since the X direction is the vibration-sensitive direction, the overall time-domain graphs of the X direction for parts 1, 2, 3, 4, and 5 were compared and analyzed to determine the optimal spindle speed.
It can be seen that the vibration trend in the X direction is consistent as the spindle speed changes. Based on the previously defined machining stages, only the vibration during the first cutting process was compared. Except for part 3, the vibration during the machining process gradually increased with increasing spindle speed. Part 3 exhibited the smallest vibration. The root mean square (RMS) value, which effectively reflects the overall vibration during the machining process, also supports this observation. It indicates that the vibration is smallest when the spindle speed is 600r/min, suggesting that spindle speeds near 600r/min are optimal. When comparing the vibration during the second cutting process, parts 1 and 2 exhibited similar and relatively small vibrations, while part 3 exhibited the largest vibration. Parts 4 and 5 had vibrations of similar magnitude, but larger than those of parts 1 and 2. With the spindle speeds of parts 1 and 2 being 570r/min and 580r/min, respectively, it further confirms that spindle speeds near 600r/min are preferable.
4. Conclusion
Through this experiment, it was found that during the first cutting process, the vibration fluctuations were significant. When the spindle speed was less than 600r/min, the vibration remained basically unchanged during the stable stage. However, when the spindle speed exceeded 600r/min, the vibration increased more notably, with the smallest vibration observed at 600r/min. During the second cutting process, the vibration was relatively stable. When the spindle speed was less than 680r/min, the vibration was small, but it increased significantly when the spindle speed exceeded 680r/min. In summary, a spindle speed near 600r/min is optimal for the first cut, and a spindle speed near 650r/min is optimal for the second cut.