Backlash analysis of a new planetary gearing with internal gear ring

Transcription

1 Journal of Chongqing University (nglish dition) [ISSN ] Vol. 9 No. September 010 Article ID: (010) To cite this article: ZHU Cai-chao XIAO Ning. Backlash analysis of a new planetary ing with internal ring [J]. J Chongqing Univ: ng d [ISSN ] 010 9(): Backlash analysis of a new planetary ing with internal ring ZHU Cai-chao 1 XIAO Ning 1 ZHANG Xiao-rong 1 State Key Laboratory of Mechanical Transmission Chongqing University Chongqing P. R. China Chongqing University of Science and Technology Chongqing 4011 P. R. China Received September 010; received in revised form 1 November 010 Abstract: Due to low carrying capacity and short life of current ring reducer we proposed a new planetary ing with internal ring which consisted of two stages. One was an involute planetary drive and the other was an N-type planetary drive with small teeth difference. The kinematic calculation was conducted based on the analysis of structural composition and working principle. Formulas of backlash calculation for the new planetary ing were derived by using probabilistic theory and the analytical model was created at the same time. Then main factors that affect the systematic backlash were introduced and the effects of manufacture error misaligments and roller bearing parameters on the distribution of backlash were presented. The influence of backlash on systematic return difference was performed based on the backlash mathematic model proposed. The results show that the backlash is a little large of which the backlash of bearing takes a greater part than two ing stages. So we presented some practical methods to reduce the systematic backlash. Keywords: planetary ing with internal ring; probabilistic theory; return difference; backlash CLC number: TH1 Document code: A 1 Introduction a The planetary ing is widely used in aerospace and industrial machinery due to its large transmission ratio high carrying capacity and impact structure. Some new and higher requirements for mechanical transmission such as larger transmission ratio lighter weight higher precision and lower backlash are presented in various industrial fields especially for high speed planetary drive systems [1-]. With the rapid development of aerospace and science technology Corresponding author ZHU Cai-chao ( 朱才朝 ): cczhu@cqu.edu.cn. XIAO Ning ( 肖宁 ): xiaoxiaoning@16.com. Funded by Key Technology R&D Program of Chongqing (No. 008AC087) and Science Technology Project of Chongqing ducation Committee (No. KJ091414). higher demands are also asked to the country s manufacture industry including accuracy of manufacture reliability strength and material among which the systematic accuracy is obviously crucial to the general level of manufacture industry [-5]. Therefore it is meaningful to conduct the research of improving the systematic precision of planetary drive. Researchers have done a lot of work on the transmission accuracy analysis of planetary drive. Blanche and Yang [6-7] discussed the influence of manufacture error of single cycloid on rotation accuracy while HIDAKAT from Yamaguchi University studied the rotation accuracy of two-stage cycloid drive structure and conducted the test. Wu et al. [8-9] investigated the return difference of RV drive. In this paper the probabilistic method [10-1] was used to perform the systematic backlash analysis of a new planetary drive with internal ring. Some practical methods were put forward to reduce the backlash which could be beneficial to the design of planetary 151

2 C. C. Zhu et al. New planetary ing with internal ring drive. Structural form and transmission principle.1 Basic structures Fig. 1 shows the structural composition of the new planetary ing with internal ring. It consists of two stages. The first stage is involute planetary drive transmission and the second is the N-type planetary drive with small teeth difference. The planet b is jointed with arm h to transmit power to the second stage. The power is inputted from shaft A and outputted from shaft B. When the ring a rotates clockwise the planet b will rotate around its own axis while rotate around the mechanism s axis. Then the external 1 of second stage will do the plane motion driven by arm h. The external will also lead the output mechanism H to rotate clockwise due to the fixed internal ring. The two stages composed of one outer ing K-V train and two parallel inner ing K-V trains [14] are coupled into a hermetic drive. Fig. shows the internal structure and Table 1 lists parameters of prototype which is designed as the structure above. Rated power of the prototype is 0.55 kw and input shaft speed is r/min. stage will do the plane motion driven by arm h. The external 1 will also lead the output mechanism H to rotate clockwise due to the fixed internal ring. The internal structure can be clearly seen in Fig. which is the breakdown drawing of prototype. Then the output motion can be obtained through output mechanism H which is clockwise. So the transmission ratio is zb z i 1 (1) z z z a 1 where z a is the tooth number of ring for the 1 st stage z b is the tooth number of planet for the 1 st stage z 1 is the tooth number of external for the nd stage and z is the tooth number of internal ring for the nd stage. 1. Output shaft;. Left box;. xternal ring; 4. Internal ring; 5. Planet ; 6. Right box; 7. Input shaft; 8. Supporting disc; 9. ccentric shaft Fig. Structure chart of the new planetary drive a: Ring ; b: Planet ; h: Arm; 1: xternal ; : Internal ; H: Output mechanism Fig. 1 Schematic of kinematics. Kinematics analysis As shown in Fig. 1 when the ring rotates clockwise the planet b will rotate around its own axis anticlockwise while rotate around the mechanism s axis. Then the external 1 of second Table 1 Prototype parameters Modification Tooth number Modulus/ coefficient mm Drive Driven Drive Driven 1 st stage nd stage Substituting the geometry parameters into q. (1) the transmission ratio can be represented by 15 J. Chongqing Univ. ng. d. [ISSN ] 010 9():

3 C. C. Zhu et al. New planetary ing with internal ring 50 6 i Fig. Internal structure and assembly chart Mathematical model of backlash The new planetary ing is a less differential drive which consists of a general involute planetary drive and a planetary drive with small teeth difference. The systematic backlash is composed of three parts which are from the first stage second stage and the bearing system. The main errors and their regularities of distribution which affects the backlash should be determined and then a probabilistic method is introduced to compute the backlash. The backlash calculated above should be converted to the output shaft and the systematic backlash can be represented by [15].1 Backlash analysis for the first stage For the first stage main factors that affect the backlash are: a) manufacture errors including reduction of tooth thickness and eccentricity which could be represented by radial integrated error and average length deviation of common normal line; b) machine tool errors and assembly errors of shaft bearing and box where we only take the offset error into account; and c) other errors which are caused by temperature and elastic deformation. These errors are ignored since they have very little influence on the systematic backlash..1.1 Average deviation of common normal line Fig. 4 shows that W is the length of normal line and is the angle between W and AB which is equal to the pressure angle. Wms Wmi and Wm are respectively the upper deviation lower deviation and tolerance among which Wm follows Gauss distribution due to its dimension error. So the equivalent value is M Wms Wmi Wm. (4) ab 1 u T T T T ab 1 u () where and T are the mean value and tolerance ab and T ab are the mean value and tolerance of geometric backlash of involute planetary drive in the first stage 1 and T 1 are the mean value and tolerance of geometric backlash of involute planetary drive with small teeth difference in the second stage and u and T u are the mean value and tolerance of geometric backlash of bearing. The total backlash caused by various factors of the new planetary ing system is T. () Fig. 4 Common normal line The equivalent value 1 D j 1 of circumferential backlash caused by the average length deviation of common normal line can be represented by j and variance J. Chongqing Univ. ng. d. [ISSN ] 010 9():

4 C. C. Zhu et al. New planetary ing with internal ring Wms Wmi j1 cos Wms Wmi D j1. 6cos.1. Gear radial integrated error (5) During the manufacturing process geometry eccentricity will appeared due to the existence of radial circle run-out. The influence of its radial component is similar to the one of the central distance deviation and obeys Gauss distribution. The equivalent value j and variance D j of backlash caused by radial circle run-out of ring are j 0 F tan D j. (6) ab equivalent value and variance T ab of return difference of the first stage are shown as ab j iah ra i T j T ab iah ra where i (8) i ah is the transmission ratio of whole system and T j is the backlash tolerance of the first stage which can be represented by T j D j. (9) i1 i.1. Offset error In Fig. 5 O 1 and O are theoretical center points while O 1 is the actual center when offset error exists. P is the theoretical mesh point while P is the actual mesh point. A is the value of offset error. Assuming the normal offset as the basic line the deviation f is symmetric with respect to the basic. While f is a size error so it follows Gauss distribution too. The equivalent value j and variance D j of backlash caused by radial circle runout of ring are j 0 f tan D j. (7) Fig. 5 Offset error where O 1 and O are theoretical center points while O 1 is the actual center when offset error exists P is the theoretical mesh point while P is the actual mesh point and A is the value of offset error..1.4 Integrated backlash for the first stage The circumferential backlash calculated above is translational backlash. However what we need is the angular backlash which is relative to the radius of reference circle. Therefore each of the backlash should be converted to the output shaft and then added up. The Therefore the return difference of output shaft caused by the first stage is T ab ab ab. (10) 154 J. Chongqing Univ. ng. d. [ISSN ] 010 9():

5 C. C. Zhu et al. New planetary ing with internal ring. Backlash analysis for the second stage..1 Inherent errors Inherent errors of include the reduction of tooth thickness and eccentricity [16]. Reduction of tooth thickness is often represented by radial feeding error r of cutting and tooth thickness allowance. Geometry eccentricity is often represented by s radial circle run-out Fr of ring or radial " integrated error F i and motion eccentric is denoted by average deviation WM of common normal line. 1) tooth thickness allowance s Assuming the normal tooth thickness as a basic line the upper deviation of tooth thickness is ss lower deviation is si and the tolerance is s (Fig. 6). Both ss and si should be minus due to backlash. s is a type of size error while submits to Gauss distribution. Fig. 6 Tooth thickness allowance The equivalent value 1 D j 1 of backlash caused by deviation of tooth thickness are ss si j1 ss si D j1. 6 j and variance (11) ) Average deviation of common normal line The equivalent value j and variance D j of circumferential backlash caused by average deviation of common normal line are Wms Wmi j cos Wms Wmi D j. 6cos (1) ) Radial circle run-out Fr of ring The equivalent value j 4 and variance D j 4 of circumferential backlash caused by radial circle run-out F r of ring are represented by j 0 4 FK r tan D j 4.. Assembly error. (1) Offset error and pitch error of holes of box are included. However only offset error is considered in this part because it has taken the pitch error of holes of box into account. Offset error is the difference of practical center distance and normal center distance which not only considers the pitch error of holes of box but also takes the errors of each component that make up the pair into account. The equivalent value j 5 and variance D j 5 of circumferential backlash caused by offset error are j5 0 fk tan (14) D j5... Integrated backlash of the second stage Based on the mathematical model of backlash the equivalent value 1 and variance T 1 of return difference of the second stage can be obtained by T i1h r1 5 i T 1 i1h r1 j i j (15) J. Chongqing Univ. ng. d. [ISSN ] 010 9():

6 C. C. Zhu et al. New planetary ing with internal ring where i 1H is the transmission ratio of the second stage and T j is the backlash tolerance of the second stage which can be represented by 5 T j D j. (16) i1 i Therefore the return difference of output shaft caused by the second stage is T (17). Backlash analysis of bearing..1 Radial clearance If the maximum value of radial clearance r max and the minimum value r min submit to Gauss distribution the equivalent value and variance can be obtained by rmax rmin j (18) rmax rmin D j. 6.. Axial parallelism error of bearing If the components of axial parallelism error of bearing with respect to the axis of plant in x and y are f x and f y respectively and obey Gauss distribution then the equivalent value and its variance can be represented by 1 B jx fxtan b (19) 1 B D jx fxtan 6 b and 1 B j y fytan b 1 B D j y fytan 6 b (0) where B is the length of bearing hole in planet in terms of mm and b is the width of in terms of mm... Backlash caused by bearing Based on the mathematical models the equivalent u value and variance T u of return difference of bearing can be represented by u ji ao i T j T u ao (1) where T j is the backlash tolerance of bearing and it can be shown as T j D j. () i1 i Therefore the return difference of output shaft caused by bearing is T u u u. () 4 Computational results The integrated backlash of the system can be obtained based on the above mathematical models. rrors affected the backlash of each stage are listed in Tables -4. Table rrors that affect backlash of the first stage Grade of accuracy Wm Ws Wi f Fr Drive 7-7-7HK Driven 7-7-7HK xternal ring Internal ring Table rrors that affect backlash of the second stage Grade of s Wm f a F r accuracy ss si Ws Wi 7-7-7HK HK J. Chongqing Univ. ng. d. [ISSN ] 010 9():

7 C. C. Zhu et al. New planetary ing with internal ring Table 4 rrors that affect backlash of bearing r r min r max f x Bearing f y According to the parameters above the backlash of each stage and the whole system can be obtained. The results are shown in Table 5. Table 5 Computational results Stage xpression symbol Computational result 1 st stage ab nd stage Bearing roller u The whole system Table 5 shows that the backlash of the new reducer is a little large of which the backlash of bearing takes a greater part and the backlash caused by the nd stage is smaller. Meanwhile the backlash caused by the 1st stage is the smallest. We presented some practical methods to reduce the backlash according to the results. 1) The crank bearing with small clearance should be chosen to improve the accuracy and reduce backlash; ) Increase the dimensional size of external ring when the requirements for design is satisfied; ) Increase the diameter of ring of the 1 st stage without affecting the radial dimension of the reducer; 4) Improve the machine accuracy and assembly accuracy of components of the whole transmission system; 5) The transmission ratio should be as large as possible and be allocated reasonably. 5 Conclusions We proposed a new planetary ing with internal ring which consists of two stages. The formulas of backlash of this kind of planetary drive are derived using probabilistic method. The prototype is manufactured based on the analysis of construction and working principle. The results show that 1) The most serious factor that affects the systematic backlash is the inherent error of which can not be inevitable. However improving the precision of machine tools and assembly accuracy can decrease the inherent error effectively; ) The influences of each stage on backlash increase from the initial stage to the final one. That is the backlash of the second stage has the greatest influence on the whole system while the backlash caused by the initial stage has a smaller influence on the system. Therefore we should try our best to control the backlash caused by the second stage; ) Transmission ratio has great influence on the backlash of transmission chains so the number of transmission stages should be reduced. The transmission ratio should be as large as possible and be allocated reasonably; 4) Backlash of train has a large relationship with ring. Therefore the diameter of ring should be as large as possible if the requirements for design are satisfied. References [1] Pomrehn LP Papalambros PY. Discrete optimal design formulations with application to train design [J]. Journal of Mechanical Design [ISSN ] (): [] Sarkar N llis R Moore TN. Backlash detection in ed mechanisms: modeling simulation and experimentation [J]. Mechanical Systems and Signal Processing [ISSN ] (): [] Walha L Fakhfakh T Haddar M. Backlash effect on dynamic analysis of a two-stage spur system [J]. Journal of Failure Analysis and Prevention [ISSN ] 006 6(): [4] Wang L. Improvement of backlash removing device: CHN TW U [P] [5] Wang L. Improvement of backlash eliminating device: CHN TW440-U [P] [6] Blanche JG Yang DCH. Cycloid drives with machining tolerances [J]. Journal of Mechanisms Transmissions and Automation in Design [ISSN ] (): [7] Yang DCH Blanche JG. Design and application guidelines for cycloid drives with machining tolerances [J]. Mechanism and Machine Theory [ISSN X] (5): [8] Wu YK Zheng JY Chen TQ et al. The analysed calculation of geometric lost motion of cycloidal [J]. Journal of Dalian Railway Institute [ISSN ] (): 8-. (In Chinese) [9] Wu YK Zheng JY Chen TQ et al. The analysis J. Chongqing Univ. ng. d. [ISSN ] 010 9():

8 C. C. Zhu et al. New planetary ing with internal ring research of the geometric lost motion of high accurate RV reducer used in robot [J]. Journal of Dalian Railway Institute [ISSN ] (): 4-7. (In Chinese). 吴永宽 郑剑云 陈天旗 等. 机器人用高精度 RV 减速机几何回差分析 [J]. 大连铁道学院学报 [ISSN ]19990():4-7. [10] Guan TM. Calculation and analysis on the return error resulting from cycloid-disk modification in the cycloid drive [J]. Modular Machine Tool & Automatic Manufacturing Technique [ISSN ] 001 4(10): (In Chinese). 关天民. 摆线针轮行星传动中修行所产生的回转误差计算与分析 [J]. 组合机床与自动化加工技术 [ISSN ]0014(10): [11] Ye QC Xu FR Wang XH. Calculating the return stagnant angle of the ing by use of probability statistic [J]. Mechanical and lectrical quipment (Marine) [ISSN ] 00 (1): (In Chinese). 叶期传 徐辅仁 王新华. 用概率统计法计算齿轮机构的回差 [J]. 机电设备 [ISSN ]00(1): [1] Xi QK. The error analysis and optimum design of cycloid ball planetary reducer [D]. Shanxi: Northwest A & F University 008. (In Chinese). 西庆坤. 摆线钢球行星减速器的误差分析及优化设计 [D]. 陕西 : 西北农林科技大学 008. [1] Chen L. Study on transmission accuracy of filtering reducer [D]. Chongqing: Chongqing University 009. (In Chinese). 陈磊. 滤波齿轮传动精度研究 [D]. 重庆 : 重庆大学 009. [14] Zhu CC Li LH Zhang L et al. Study on double-crank reducer with external board [J]. Transactions of the Chinese Society for Agricultural Machinery [ISSN ] 008 9(8): (In Chinese). 朱才朝 黎利华 张磊等. 双曲柄外齿环板减速传动的研究 [J]. 农业机械学报 [ISSN ]0089(8): [15] Li CN Liu JY Sun T. Study on transmission precision of cycloidal pin in K-V planetary drives [J]. Chinese Journal of Mechanical ngineering [ISSN ] 001 7(4): (In Chinese). 李充宁 刘继岩 孙涛.K-V 型行星传动中摆线针轮啮合的传动精度研究 [J]. 机械工程学报 [ISSN ]0017(4): [16] Zhou AQ. Computing backlash for the transmitting mechanism in which the difference in the number of teeth of the meshing s is small [J]. Acta Armamentarii: [ISSN ] (4): 4-9. dited by XU Jing-yuan 158 J. Chongqing Univ. ng. d. [ISSN ] 010 9():