TECHNICAL PAPER Development of Super-low Friction Torque Technology for Tapered Roller Bearing H. MATSUYAMA H. DODORO K. OGINO H. OHSHIMA H. CHIBA K. TODA To achieve high efficiency in rear axle differentials, reduction in friction torque of tapered roller bearings which support drive pinions was investigated. As a result, effects of design factors and oil flow inside the bearing on friction torque were clarified. Friction torque of a newly developed bearing which was optimized considering various performance characteristics was 75% lower than that of a conventional bearing. This is equivalent to a friction torque reduction of about 8% compared to a standard bearing. This reduction is expected to have a remarkable effect on vehicle fuel efficiency. Key Words: tapered roller bearing, friction torque, rear axle differential, fuel efficiency 1. Introduction Automobiles have a tremendous effect on the global environment. It is imperative to improve their fuel efficiency and to reduce the amount of CO emission. Therefore, in addition to high reliability and low cost, more compactness, lighter weight and higher efficiency are demanded for automotive components. This study, by focusing on the improvement in efficiency of automotive rear axle differentials, was conducted to reduce the friction torque generated in the bearing which supports the drive pinion. In general, tapered roller bearings, which have both high load carrying capacity and high stiffness, are used for the differentials. However, some automotive manufacturers have recently considered the use of ball bearings for reduction of friction torque 1). Since life, stiffness and static safety of ball bearings are inferior as compared to comparable-sized tapered roller bearings, much larger size ball bearings are required to provide the equivalent performance of tapered roller bearings ). Therefore, there is a demand to develop a new low friction torque tapered roller bearing while maintaining long life and high stiffness. In this paper, experimental and calculation results of factors that cause torque in tapered roller bearings supporting drive pinions are shown with their estimated contribution ratios. Studies are also conducted for clarifying the effects of internal bearing geometry (roller effective length, number of rollers, contact angle, pitch circle diameter of rollers and raceway crowning profile) and lubricant flow through a bearing on the torque, and, based on these insights, a new tapered roller bearing with super-low friction torque has been developed. Results of this study are shown hereinafter.. Factors Causing Torque in Tapered Roller Bearing Generally, friction of a tapered roller bearing is caused by the following four factors. (A) Viscous rolling resistance between rollers and raceways (B) Sliding resistance between rollers and the rib (C) Agitating resistance of lubricant (D) Sliding resistance between rollers and the cage In previous studies 3)-6), the effects of (C) and (D) have often been ignored because the impact on torque is relatively small compared to (A) and (B). However, in the case of differentials where high viscosity gear oil is used, the effect of (C) hardly seems negligible. Thus, experiments were conducted to clarify factors, and their contributions, that cause torque in the pinion support bearing. Initially, using a special differential carrier made of transparent acrylic resin, oil flow around bearings was observed. Figure 1 shows a typical structure of a rear axle differential. As in a production vehicle, the carrier was filled with gear oil, which was supplied to the pinion support tapered roller bearings by the motion of ring gear. As the result of observation, it was found that both the head and the tail bearings were perfectly satisfied with the oil at the pinion speed of 6~9 min 1 (corresponding to ~3 km/h vehicle speed) or more. This implies that analysis of the bearing torque caused by oil agitation, as shown in (C), should be quantitatively conducted at speeds encountered under practical operating conditions in the field. Koyo Engineering Journal English Edition No.167E (5)
Ring gear Head bearing Drive pinion Fig. 1 Rear axle differential Oil flow Tail bearing Next, the effect of oil flow through single row tapered roller bearing with the bore diameter of 3mm and the outside diameter of 7mm on the torque was investigated. A test was carried out under the condition shown in Table 1, which represents the new ECE driving mode (93/116/EC) 7). Figure shows the schematic diagram of test equipment. The test bearing was axially loaded and then the inner ring was rotated at the fixed speed. The oil level was maintained constant by replenishing the same amount as had passed through the bearing. The torque was obtained by measuring the moment acting on the outer ring that is supported by an air bearing. In this test, torque measurement began with a full oil flow condition as shown in Fig. (a stable condition wherein the bearing back face is filled with the oil and the oil level is kept at mm). The measurement was then repeated at successive reduced oil flow rates. Rotational speed Table 1 Test conditions min 1 Axial load kn Lubricant Gear oil 85W 9 Oil temperature 5; Lubrication Circulating (full oil flow) Oil mm (Oil height) Load Air bearing Load cell Test bearing Test results are shown in Fig. 3. The horizontal axis of the figure shows the ratio of oil flow with the full oil flow as 1, whereas the vertical axis shows the torque ratio with the torque at full oil flow set at 1. From Fig. 3, it can be seen that the torque decreases with a decrease in oil flow rate. At the minimum oil flow.1, that can avoid seizure, the torque ratio is about. If it is assumed that the reduction of torque with the reduced oil flow rate was due to reduction of agitating resistance, this implies that the contribution of oil agitating resistance to the torque at full oil flow is about 3%. Based on this result, contribution of each factor related to the torque of tapered roller bearing supporting the drive pinion under the ECE driving mode condition was estimated. Previous studies on this subject 3)-6) have shown that the bearing torque is the sum of factors (A) through (D), and that (D) can be ignored. And, as mentioned above, the contribution of factor (C) was regarded as 3%. The torque contribution of factor (B) was theoretically calculated using the friction coefficient between the roller end face and the rib, which is estimated from the oil film parameter 8). As shown in Fig., the result indicate that the greatest contribution of 65% was obtained from the viscous rolling resistance between the rollers and the raceways, followed by 3% from the agitating resistance of oil, while the contribution of sliding resistance between the roller end and the rib was as small as 5%. This is due to the formation of a sufficient EHD film between the roller end and the rib under high-viscosity gear oil lubrication. Based on the results shown in Fig., in this study, to further reduce the torque of existing low friction torque 9), 1) tapered roller bearings (LFT bearing) in commercial production, following studies were conducted; a optimization of internal design factors for further reducing viscous rolling resistance and s improvement of oil flow for reducing agitating resistance..5.. 1 Ratio of oil flow rate Fig. 3 Effect of oil flow rate on friction torque Motor Fig. Schematic diagram of test equipment Koyo Engineering Journal English Edition No.167E (5) 3
3. Reduction of Viscous Rolling Resistance by Optimization of Internal Design In order to reduce the viscous rolling resistance, it is effective to reduce the rolling contact area by decreasing the number of rollers and the roller effective length. However, these changes may result in reducing the load carrying capacity and reducing the bearing stiffness. Therefore, tests were conducted to experimentally determine the effects of internal design factors on torque under full oil flow condition, and the internal design factors were optimized while balancing the performance. For these tests, single row tapered roller bearings with bore diameter from 3mm to 55mm and outside diameter from 7mm to 1mm were used. The evaluated design factors are shown in Table and Fig. 5. In all tests, the axial load was 6 kn and the rotational speeds were 1, and 3 min 1 while all other test conditions are shown in Table 1. Symbols LWR z a dm RCi DW Agitating resistance 3% RCo dm Table Evaluated design factors Design factors Roller effective length Number of rollers Contact angle (outer raceway half angle) Pitch circle diameter of rollers Crowning radius of outer raceway Crowning radius of inner raceway Mean roller diameter DW LWR a Sliding resistance 5% Viscous rolling resistance 65% Fig. Factors contributing to friction torque of tapered roller bearing used for drive pinion support in rear axle differential Raceway profile RCo or RCi Fig. 5 Bearing internal geometry The relationships between the design factors and the torque are shown in Fig. 6. The following relationships are shown in the figures. 1) The torque decreases when the roller effective length is decreased. ) The torque decreases when the number of rollers is reduced. 3) The torque decreases when the contact angle is increased. ) The torque decreases when the pitch circle diameter of rollers is decreased. 5) The torque decreases when crowning radius of the inner or outer raceway is decreased. The torque decreases with decreasing the crowning radius, because the contact length between the roller and the raceway is shortened. Also, in each figure, a dashed line was added to show the values 8) calculated based on EHD analysis and experimental data under bath lubrication. As far as roller length, number of rollers, pitch circle diameters of rollers and raceway crowning radius are concerned, the test results are consistent with the calculation results. However, reduction of the torque due to the increase of contact angle is not consistent with the calculation results, which is considered to be due to an improvement in pumping performance. These test results are summarized in Table 3, where the mean roller diameter, DW, is based on previous studies 3). In addition, this summary also shows the effects of each design factors on the dynamic capacity and the stiffness of the bearing, which were calculated with the basic dynamic capacity and the elastic deformation 11). With upward and downward arrows indicating the increase and the decrease, respectively, the arrows in this table indicate how each design factor should be changed to improve the bearing performance. For higher efficiency, that is, to reduce the torque, the contact angle, a, should be increased, while the number of rollers, z, and the roller effective length, LWR, should be decreased. However, the increase of a reduces the dynamic capacity, and the decrease of z and LWR would make both the dynamic capacity and the stiffness lower. While DW has greater effects on the dynamic capacity than z or LWR, the effect of DW on the torque would be far smaller than z or LWR. Therefore, in order to reduce z which most effectively reduces the torque, an increased DW is necessary to offset the reduction of load carrying capacity due to reduced z. This seemed to be the optimum design that can be effective without changing the boundary dimensions of the bearing. Table 3 Optimum design guidelines Performance LWR z a dm RCo RCi DW High efficiency High capacity High stiffness Optimum design Koyo Engineering Journal English Edition No.167E (5)
1. 1. 1.3 1... 1 1 1 16 18. 8 1 16 LWR, mm z, pcs. (a) Roller effective length (b) Number of rollers 1 15 5 3 a, deg (c) Contact angle 5 3 1 5 6 7 8 9 1 3 1 3 dm, mm RCo, mm RCi, mm (d) Pitch circle diameter of rollers (e) Crowning radius of outer raceway (f) Crowning radius of inner raceway Fig. 6 Effects of bearing design factors on friction torque ( : 1 min 1, : min 1, : 3 min 1, : Experimental mean line, : Calculated value). Reduction of Agitating Resistance by Oil Flow Improvement As shown in Fig. 3, it is possible to reduce the agitating resistance by reducing the oil flow through the bearing. Therefore, the effect of oil flow control on the torque was investigated by changing the oil flow rate using interruption rings attached on the bearing. The test bearing was a single row tapered roller bearing of 3mm bore diameter and 7mm outside diameter and four kinds of attachments as shown in Fig. 7 were used. Tests were conducted under a constant axial load of kn and rotational speeds of 5, 1, 1 6 and min 1. All other conditions were identical to those shown in Table 1. The relationship between rotational speeds and oil flow rate is shown in Fig. 8. The oil flow rates of all the bearings with the attachments were less than those of the normal bearing. The effect of Type A was largest. The oil flow rate of Type A at min 1 is about.5 L/min, and is reduced by over 9% from that of the normal bearing. The effect of Type B was smallest. Subsequently, the relationship between the rotational speed and the torque is shown in Fig. 9. The torque of Type A and Type B, where the oil inlet side is restricted, is lower than that of the normal bearing. Especially, at the speed of min 1, the torque of Type A is.5 N m which is 3% less than the normal bearing. On the other hand, the torque of Type C and D, where the oil exit side is restricted, is higher than that of the normal bearing. Oil inlet side (b) Type A (a) Normal Oil exit side (c) Type B (d) Type C (e) Type D Fig. 7 Test bearings with attachment for restricting oil flow Koyo Engineering Journal English Edition No.167E (5) 5
Oil flow rate, L/min 1 8 6 Normal Type A Type B Type C Type D (a) Normal (b) Type A Fig. 1 Observation of oil flow in bearing Friction torque, N m 1 3 1.. Fig. 8 Relationship between rotational speed and oil flow rate Normal Type A Type B Type C Type D. 1 3 Fig. 9 Relationship between rotational speed and friction torque In Type C and Type D, the oil flow decreased, but the torque increased inversely. It is theorized that the oil stagnation inside the bearing, from the restriction at the exit side of the bearing, increased agitating resistance. Therefore, to reduce the agitating resistance, it is important to minimize oil stagnation by rapidly discharging the oil inside the bearing 1). For Type A and Type B, there is a difference in their effects on oil flow rates and their effects on friction torque. It is theorized that the clearance between the cage and the front face rib of the inner ring has a great influence on the oil inflow to the bearing, and so it is more effective for the reduction in agitating resistance to decrease this clearance. Figure 1 shows observation result of the oil flow in the bearing through a transparent acrylic-resin outer ring in the similar method as previously reported 1). While the normal bearing showed cavitations in the lubricant indicating that violent agitation was occurring therein, no such cavitations were observed on Type A with decreased clearance between the cage and the front face rib of the inner ring indicating smooth oil flow free from violent agitation. From this observation, the agitating resistance of Type A is considered to be smaller than that of the normal bearing. As shown in Fig. 6 (c), it is theorized that the decrease in torque due to increased contact angle is resulting from a decrease in oil stagnation in the bearing, by improvement of the bearing pumping efficiency 13). 5. Performance of Super-low Friction Torque Tapered Roller Bearing A super-low friction torque tapered roller bearing has been developed based on the test results described in this report. Comparison of the developed bearing with a conventional one is shown in Fig. 11. Compared with the conventional bearing, the developed one features larger contact angle, smaller number of rollers, shorter roller length, larger roller diameter and special crowning profile on the inner and outer raceways. Moreover, the cage bore diameter is smaller to reduce the clearance between the cage and the inner front face rib. Comparison tests of the bearing torque and the oil flow rate were conducted for the developed bearing, conventional (LFT bearing) and standard bearings under the test conditions as shown in Table 1. Relationship between the rotational speed and the torque is shown in Fig.1 and that between the rotational speed and the oil flow rate is shown in Fig. 13. As can be seen from Fig. 1, the torque of the developed bearing is N m at min 1, which is 5% lower than the conventional bearing, or 6% lower than the standard bearing. Also, as seen in Fig. 13, the oil flow rate of the developed bearing is smaller than that of the other bearings. But, at 3 min 1, this rate is 3.6 L/min, which is more than 3% of the oil flow rate observed on the conventional bearing. Lower oil flow rates seem to have caused somewhat higher temperature rise on the developed bearing than the conventional one as measured on the outer ring. However, the temperature rise is higher by not more than ;. So, there is no danger of seizure for the developed bearing. Next, further lower friction torque tapered roller bearing has been developed by applying Koyo's original material and heat treatment technology 1)-16) that allow bearings to be smaller in addition to the new technology reported here as shown in (a) Developed (b) Conventional Fig. 11 Comparison of developed bearing with conventional bearing (bore diameter: 35mm, outside diameter: 89mm, width: 38mm) 6 Koyo Engineering Journal English Edition No.167E (5)
Fig. 1. The developed bearing showed as a low friction torque as N m at min 1 which represents 75% reduction from the conventional bearing, or 8% reduction from the standard bearing. Furthermore, this figure shows measured torque data of a ball bearing designed to have the same calculated life as the tapered roller bearing. Compared with this ball bearing, the developed tapered roller bearing shows % lower torque. As mentioned above, putting together all these technologies for reducing bearing torque, it is now possible to supply a super-low friction torque tapered roller bearings which torque is lower than a ball bearing for the market. Since, in some differentials, more than 5% of the friction loss could be attributed to the drive pinion bearings, super-low friction torque of the new drive pinion bearings is expected to improve vehicle fuel efficiency by as much as % or more with the ECE driving mode. Friction torque, N m 3..5. 1.5.5 Standard Conventional Developed. 1 3 Fig. 1 Relationship between rotational speed and friction torque Oil flow rate, L/min 1 1 8 6 Standard Conventional Developed 1 3 Fig. 13 Relationship between rotational speed and oil flow rate Friction torque, N m 5 3 1 1 3 6. Conclusions Conventional (5 115 5) Ball bearing (5 135 55) Developed (5 18 3.5) Fig. 1 Super-low friction torque performance of developed bearing with downsizing (Dimensions in parentheses are ID OD Width.) To reduce the torque of tapered roller bearings that support automotive differential pinions, factors contributing the torque were analyzed and the reduction of both viscous rolling resistance and agitating resistance, whose torque contributions are high, were investigated. Specifically, the effect of internal design on the viscous rolling resistance and that of the oil flow control on agitating resistance of oil was clarified. As a result, following knowledge was obtained. 1) Under the representative conditions of the ECE driving mode, the torque of tapered roller bearing is made up of 65% viscous rolling resistance between the rollers and the raceways, and 3% agitating resistance of the oil. ) The design concepts for low friction torque tapered roller bearings with balanced performance are as follows: To decrease the viscous rolling resistance, the rolling contact area should be reduced by decreasing the number of rollers, roller effective length and the crowning radius of raceways. To decrease the agitating resistance, the oil inflow into the bearing should be controlled by reducing the clearance between the cage and the inner front face rib. Furthermore, oil stagnation should be decreased by increasing the contact angle to enhance the pumping efficiency in the bearing. To offset the decline in the dynamic capacity due to the reduction in the number of rollers, the mean roller diameter, which has less impact on the torque, should be increased. 3) The super-low friction torque tapered roller bearing, which was developed on the basis of this study, has achieved a friction torque reduction of 5% compared to the couventional bearing of the same size. Furthermore, another new tapered roller bearing incorporating special heat treatment technology to reduce the bearing size in addition to the low friction torque concepts has achieved a 75% torque reduction. These technologies are expected to remarkably contribute to improved fuel efficiency of vehicle. Koyo Engineering Journal English Edition No.167E (5) 7
References 1) D. Sprindler and G. V. Petery: SAE Technical Paper, 3-1-373 (3). ) K. Yoshimura, K. Ohshima, K. Kondo, S. Ashida, H. Watanabe and M. Kojima: Journal of Society of Automotive Engineers of Japan, 58, 9 () 68. 3) D. C. Witte: ASLE Trans., 16, 1 (1973) 61. ) S. Aihara: Trans. ASME, Journal of Tribology, 19, 3 (1987) 71. 5) R. S. Zhou and M. R. Hoeprich: Trans. ASME, Journal of Tribology, 113, 3 (1991) 59. 6) H. Matsuyama, S. Kamamoto and K. Asano: SAE Technical Paper, 989 (1998). 7) DCBACK. CO., LTD., Driving Mode System, http://www.dcback.com/bbs/data/test/1676788/introdu ction_for_company.doc. 8) H. Matsuyama and S. Kamamoto: Koyo Engineering Journal, 159E (1) 53. 9) M. Takeuchi: Koyo Engineering Journal, 17 (1985) 5. 1) Y. Asai and H. Ohshima: Koyo Engineering Journal, 13 (1993) 3. 11) T. A. Harris: ROLLING BEARING ANALYSIS, 3rd Ed., John Willey & Sons (1991) 37 & 7. 1) K. Toda, M. Shibata and T. Hoshino: JAST Trans.,, (1997) 38. 13) T. Ohmori, J. Okamoto and T. Wakabayashi: Proceedings of JAST Tribology Conference, Kitakyusyu (1996) 9. 1) K. Toda, T. Mikami and T. M. Johns: SAE Technical Paper, 9171 (199). 15) M. Gotoh: Koyo Engineering Journal, 165E () 31. 16) K. Kizawa and T. Mikami: Proceedings of JAST Tribology Conference, Niigata (3) 5. H. MATSUYAMA * H. DODORO ** K. OGINO *** H. OHSHIMA **** H. CHIBA * K. TODA * * Core Technology Research & Development Department, Research & Development Center ** Chubu Technical Center, Bearing Business Operations Headquarters *** Analysis Engineering Department, Bearing Business Operations Headquarters ****Automotive Bearing Engineering Department, Bearing Business Operations Headquarters 8 Koyo Engineering Journal English Edition No.167E (5)