Study on Mechanism of Impact Noise on Steering Gear While Turning Steering Wheel in Opposite Directions Jeong-Tae Kim 1 ; Jong Wha Lee 2 ; Sun Mok Lee 3 ; Taewhwi Lee 4 ; Woong-Gi Kim 5 1 Hyundai Mobis, South Korea 2 Hyundai Mobis, South Korea 3 Hyundai Motor Company, South Korea 4 Psylogic, South Korea 5 Virtual Motion, South Korea ABSTRACT This study is focused on the cause of the clanking noise which called "Tuk" in a vehicle. The noise was generated from a steering gear system under high load conditions while a steering wheel was turning in the opposite direction. In order to identify the mechanism of the noise, both experimental and simulational studies were performed on a steering gear system in lab-testing conditions. A simulation model was constructed based on modal testing and deflection data measured at several points in operating conditions. The detailed behavior of each component such as a rack bar, a yoke and a housing was able to be investigated with the help of the transient analysis. As a result of the testing and the simulation, it was concluded that a vibration was caused by the collision of a rack bar and a pinion gear. The vibration which started from the gear interface was transferred to the neighboring parts and the noise was radiated mainly at the housing. The impact phenomenon was additionally confirmed with the measurement of transmission errors between the gears. Further studies to suppress the noise have been successfully performed with the help of the analysis model obtained from this study. The effectiveness of final countermeasures was also verified with testing. Keywords: Steering Gear, Impact Noise, CAE I-INCE Classification of Subjects Number(s): 76.9 1. INTRODUCTION The most significant noises that cause customer s complaints in the steering gear of an electronically assisted power steering (EPS) system are rattle and clanking noise. The rattle is periodic noise due to the impact between gear teeth when a vehicle is driven on an unpaved road, and the researches on the gear rattle including other applications have been carried out for many years (1-4). The clanking noise, called Tuk, can be heard inside a cabin when turning the steering wheel in the opposite direction in an engine idle condition. The noise is assumed to be caused by the impact between internal moving parts in the steering gear system. The clanking noise has been major issue for many years, however it is challenging to clearly identify the source because of numerous contacts between the parts such as a rack bar, a pinion shaft, a yoke and etc. For this reason, the research on the mechanism of the clank noise was rarely done, and the reduction of noise was made mainly rely on testing based on trial and error. 1 jtkim@mobis.co.kr 2 jongwha@mobis.co.kr 3 sunmk@hyundai.com 4 thlee@psylogic.co.kr 5 woonggikim@virtualmotion.co.kr 7314
The purpose of this study is to identify the mechanism of a clanking noise in the steering gear of an EPS system in order to find effective method to reduce the noise. Both testing and simulation were performed on a steering gear system in lab-testing conditions. First, the motion of rack bar and yoke body was measured while the steering wheel was turning left to right and right to left, repetitively. Noise and vibrations at some points on the housing were also monitored to see the relationship with the motion measured in operating condition. Based on the testing results, a simulation model was built to observe detailed behavior of each part and the mechanism of the collision between each part were found. Finally, the impact mechanism was additionally confirmed with the help of transmission errors between the rack and pinion gears. 2. MEASUREMENT OF BEHAVIOR 2.1 Measurement of Displacement Lab testing was carried out to observe detailed motion of moving parts such as a pinion shaft, a rack bar, and a yoke body as shown in Figure 1. Figure 2 shows the equipment of the lab testing where external loads were exerted to the end of the tie rod to simulate tire reaction force of a real vehicle during a steering operation. The angle of the pinion shaft and the displacement of the rack bar were recorded using encoder, and displacements at several major points were measured with LVDT s. The sensors were installed with additional fixtures and their detailed locations including the directions were shown in Figure 3. The load cell was also inserted in the left and right tie rods of the steering gear system to obtain the external loads transmitted from the motor for tire reaction forces. Figure 1 Structure of steering gear system. Figure 2 Lab testing equipment. Figure 3 Sensor location. 7315
The testing was performed for the steering gear system while the steering wheel was repetitively turning in the opposite directions from left to right and right to left. As a result of the testing, the motion with time is plotted in Figure 4. The greatest clanking noise was generated between 11.25 ~ 11.3 sec and a high level of the acceleration on the housing appeared at the same time. The yoke and the rack bar also showed abrupt change in displacement at that time interval. Before the measurements were performed, the noise was supposed to occur at the instance of changing directions by a steering wheel. The noise, however, was generated when the steering wheel is passing by right before the center location where the external load is close to zero. Figure 4 Measured angle and deflection of steering gear. The yoke gap in Figure 1 is defined as the absolute difference between the maximum and minimum displacement in the X direction. The yoke gap is designed to ensure low steering efforts under a high load condition but it causes unwanted motion of the yoke and its surrounding parts. The yoke was stayed at its -X extreme position by the amount of the gap most of the steering operations. It can be seen that the yoke moved toward the center of the pinion shaft in +X direction and the displacement of the yoke was decreased abruptly right before the peak of the vibration. The displacement of the rack bar at the L1 position was closely related with the X displacement of the yoke, and hence it is assumed that the yoke was in contact with the rack bar in the X direction. The rack bar was bended during the steering operation and its deformed shape was presented in Figure 5 when the maximum deflection reached. The figure shows that the amount of deflection was increased with the increment of the yoke gap. Figure 5 Deformation of rack bar 2.2 Investigation on Noise In order to investigate noise characteristics during the clanking noise event, sound pressure levels and accelerations were measured at the locations shown in Figure 6. 7316
Figure 6 Noise and vibration measurement As shown in Figure 7, the acceleration of the rack bar in the X direction was the greatest compared to those of the Y and Z directions. Consequently, it was considered that the clank noise was mainly originating from an impact in the X direction. The peak of the acceleration on the center of housing followed the peak of the acceleration on the rack bar or on the housing near the yoke after about 0.003 sec. It can be seen that the impact took place around the yoke or the rack bar and the vibration was transferred to the center of the housing. Figure 7 Acceleration and Sound Pressure The measured noise was analyzed in both time and frequency domain using wavelet transformation as shown in Figure 8. Figure 8 Noise spectrum The frequency at the peak of the noise was closely related to the resonant frequency of the 7317
housing. The relationship can be seen more clearly if the noise and vibration spectrums which measured at each component were compared. The vibration spectrum is shown in Figure 10. The instant timing at the vibration peak of the housing differed from those of the rack bar to the pinion shaft. Figure 9 Vibration spectrum Figure 10 Sound visualization The noise was also visualized using a sound intensity meter in order to find where the noise was mainly radiated on the housing surface. The noise was radiated from the center of the housing as shown in Figure 10. As a result of the testing, it can be concluded that a vibration occurred near the rack and pinion gears transferred to the housing and it leads to the noise radiated from the center of the housing. 7318
3. ANALYSIS 3.1 Modeling In order to investigate detailed motion of the rack bar, pinion shaft, the yoke and etc with time, transient analysis on the steering gear system were performed using commercial S/W, DAFUL. The analysis model was constructed as shown in Figure 11. The rack bar was modeled with beam elements to realize the bending as observed in the testing. Mount bushes and O-rings were modeled as springs with 6 degree of freedom. The rack bush and the P-bush were also modeled with spring elements and contact conditions were applied at the surface adjacent to the rack bar. Furthermore, contact conditions were applied at the interfaces of other parts including the yoke as shown in Figure 12. Through the whole analysis process, the impact phenomenon of each part can be described effectively. Figure 11 Analysis model. Figure 12 Modeling of rack and pinion gear. 3.2 Load and Boundary Conditions Figure 13 Load and boundary conditions 7319
For load conditions of the analysis model, measured data from lab testing were used. The loading data measured at the tie rods showed that the forces increased up to 2500N with enlarging angle of the steering wheel, and decreased as the steering wheel returned to its neutral position. In the analysis, external forces were applied at the tie rods as shown in Figure 13. Fixed boundary conditions were applied to the mounding of the housing and the hinge connectors were added to simulate lab testing conditions. Also, the measured angle was input to the pinion shaft up to 0.6 sec which corresponds to almost one cycle of the steering position from neutral to right-hand, right-hand to left-hand, and left-hand to neutral position. Basically, the time step was set to 0.005 sec during the analysis but it was refined to 6.6x10-6 sec for the time range from 0.27 to 0.3 sec where the impact was expected from the testing results. 3.3 Results The transient analysis was performed with the input data obtained from the testing referred to the previous section 2.1. The Figure 14 shows a comparison between the displacements of the testing and those of the analysis at the yoke and the rack bar. Generally, there were similar behaviors between the testing and analysis. As shown in the case of the testing, the analysis results also showed that the yoke stayed in contact with the plug for most of the steering, especially under the high load condition. The yoke moved towards the rack bar and then there were sudden change in displacement at 0.2812 sec when the steering wheel is turned from right-hand to neutral position or left-hand to neutral position. a) b) Figure 14 Displacements of testing and analysis; a) yoke, b) rack bar a) b) c) Figure 15 Deformed shape of rack bar(50 times magnified); a) 0.2762 sec, b) 0.2796 sec, c) 0.2812 sec. 7320
The scaled deflection shape of the rack bar is presented in Figure 15 to illustrate how the rack bar deformed and where the impact occurred with time. Figure 14a shows the case when the steering wheel angle was decreasing from 41 deg to 10 deg. The yoke maintains the position in contact with the plug because the external load transferred from the rack bar is greater than the spring force which pushed the yoke in the +X direction. The rack bar undergoes bending deformation similar to 2 nd bending mode shape because the rack bar was supported by the yoke in the left and the rack bush in the right. As the steering wheel angle decreased less than 10 deg, the external load decreased gradually. This lead to the yoke motion toward the pinion gear because the spring force became greater than the external load transferred from the rack bar as shown in Figure 15b. The left-hand side of the rack bar also moved toward the pinion gear because the rack bar was always in contact with the yoke. The motion of the rack bar was more accelerated because external load was continuously decreasing and the restoring force of the rack bar helped the rack bar motion. Finally, due to the accelerated rack bar, the teeth of the rack bar collided with the teeth of the pinion gear as shown in Figure 15c. The impact between the rack and pinion gears is described in more detail in Figure 16 where the red arrows indicated contact forces. Figure 16 Contact force of the rack and pinion gear; a) before collision, b) after collision. Before the collision, as shown in Figure 16a the right teeth of the pinion gear were contact with the left teeth of the rack gear. After the collision, the additional contact between the left teeth of the pinion gear and the right teeth of the rack gear took place as shown in Figure 16b,. a) b) Figure 17 Velocity results; a) velocity of yoke bar with time, b) velocity distribution of rack bar before collision. The vibration induced by the impact is known to be proportional to the collision velocity. The yoke was in contact with the rack bar during the collision and it could be clearly defined compared to the velocity of the rack bar. That s why the yoke velocity could be calculated in Figure 17 in order to evaluate the strength related to the source of vibration. The velocity of the yoke increased consistently until it reached its maximum speed of 40 mm/sec. after the collision, the velocity of the 7321
yoke dropped rapidly and then kept around 0 mm/sec. The velocity distribution of the rack bar also plotted in Figure 17. This graph indicates that the velocity profile was nearly linear and the velocity at the rack bush was almost zero. Consequently, it can be seen that the rack bar was moving like a rigid body in rotational motion hinged at the rack bush right before the collision occurred. The vibrational characteristics of the analysis were plotted in Figure 18. The acceleration of the rack bar and the housing in Figure 18 showed similar tendency with time compared with the testing results in Figure 7. A large vibration was also observed after the rack bar was collided with the pinion shaft. Figure 19 shows how the vibration generated at the interface of rack and pinion gear was propagated with time. Before 0.2811 sec, there was no vibration on the housing. The vibration started at 0.2812 sec from rack and pinion gear and the vibrational wave traveled to the opposite side of the housing. After the reflection wave started from 0.2813 sec, a standing wave started to be built up in the housing. a) b) Figure 18 Analysis results; a) acceleration at rack bar, b) acceleration at housing. Figure 19 Acceleration contour on housing with time. 7322
4. TRANSMISSON ERROR In order to confirm the impact between the rack and pinion gear when the clanking noise occurred, transmission error of the gear was examined using the data obtained in the section 2.1. Transmission error, TE of the rack and pinion gear can be defined as follows. TE L r (1) Where θ is pinion angle, r is radius of pitch circle, and L is translational displacement of rack bar. Transmission error and its derivative with respect to time (DTE) are presented in Figure 20. From this figure, it is seen that the transmission error increased with increasing external load. This is because the distance between the gear axes increased depending on the external load. The instant timing of clank noise was exactly coincident with the timing of peak of DTE. a) b) Figure 20 Transmission error of rack and pinion gears; a) transmission error and displacement, b) time derivative of pinion and rack bar during collision. The speed of the rack bar decreased while the rotational speed of the pinion shaft increased before the 1 st peak of the acceleration between 1.299 sec and 1.3 sec. After the 1 st peak of the vibration, the speed of rack bar turned to increase with time and the rotational speed of the pinion shaft started to decrease. This implies that the distance between the gear axes altered with time after the 1 st peak of the vibration. From the analysis on the transmission error, it is reconfirmed that the clank noise was induced by the collision of the rack bar and the pinion shaft. 5. CONCLUSIONS In order to identify the clank noise in the steering gear system, both lab testing and transient analysis were performed. As a result, the detailed motion of the steering gear system was investigated and the following facts were found. 1) The cause of the clank noise was identified as a low speed collision between the rack bar and pinion shaft. 2) The noise was generated when the steering wheel was passing the center position, not when 7323
the steering wheel was changing a rotational direction. 3) The mechanism of noise generation was as follows. - The rack bar underwent bending deformation and the yoke was in contact with the plug when the steering wheel was turned left or right. - As the steering rotated to the center, yoke and the left of the rack bar start to move toward the pinion and accelerated due to the restoring force and the decreasing external force. -The moving rack bar collided with the pinion gear teeth at a relatively low speed. -Vibration was generated from the gear interface as a result of the impact -The vibration was transferred to the housing and a clanking noise was radiated from the housing. 4) The analysis model to predict the clank noise was built and it showed close correlation with the test results. By applying the analysis model obtained from this study, the following further studies to reduce the clanking noise have been successfully performed. As a result of a parameter study several improved designs were proposed. Finally their effectiveness was verified with additional testing. REFERENCES 1. Choi S, Kim R, Choi K. Study on the MDPS Vibration Analysis, KSAE Annual Conference & Exhibition 2009;2009.11:2274-2283 2. Fernholz C. A Simplified Approach to Quantifying Gear Rattle Noise Using Envelope Analysis, SAE Technical Paper 2011;2011-01-1584. 3. da Silva J, Zanini J. Design of Experiments Application (DOE) to Prevent Mechanical Noise in Power Rack & Pinion Steering Systems, SAE Technical Paper 2004;2004-01-3377. 4. Ognjanović1 M, Kostić SĆ. Gear Unit Housing Effect on the Noise Generation Caused by Gear Teeth Impacts, Journal of Mechanical Engineering 2012;58 (5):327-337. 5. Yufang W, Zhongfang T. Sound radiated from the impact of two cylinders, Journal of Sound and Vibration 1992;159 (2):295-303. 6. Mehraby1 K, Beheshti HK, Poursina M, Impact noise radiated by collision of two spheres: Comparison between numerical simulations, experiments and analytical results, Journal of Mechanical Science and Technology 2011;25 (7):1675-1685. 7324