KINEMATICS, DYNAMICS AND MECHANICAL EFFICIENCY OF A CARDAN JOINT WITH MANUFACTURING TOLERANCES - Part II

E. Pennestr`, L. Vita, P.P. Valentini

Abstract: This second part of the paper summarizes the methodology for dynamic and mechanical efciency analysis of a Cardan joint. The numerical results have been also experimentally validated. Keywords: Mechanical transmissions, Cardan joint, Dual numbers, mechanical efciency. 1. INTRODUCTION

Although the kinematics of the Cardan joint seems extensively studied, the contributions on models for internal forces analysis are scarce. The rst studies on the dynamics of the Cardan joint appear around 1930-1940 and may have been prompted by the numerous joint failures observed in the drivelines of automobiles. These studies hinted the presence of rocking torques or secondary couples, with direction orthogonal both to input and output shafts, as causes of the failures. To the best of our knowledge, the most detailed approach to the dynamic analysis of Cardan joints is due to F. Freudenstein and his coworkers (e.g. [1, 2]. In a series of papers, making use of dual vectors, the equations of motion have beed deduced in symbolic form. The main steps of their deductions are herein outlined. Regarding mechanical efciency of Cardan joints, likely the rst scientic contribution is due to Morecki [3]. Such a model, based on a simplied static analysis and including the losses in the yoke bearings only, has been veried by E. Pennestr` et al. with a different analytical approach and rened by including also the losses in the xed bearings [5, 6]. The results have been plotted in a design chart which allows to compute the efciency as a function of the angle between input-output shaft axes . The inuence on energy losses of angular speed, input and output shafts angular misalignement, have been investigated by E. Pennestr` et al. [7]. The numerical results have been also experimentally validated [10]. In this second part the main ndings of the different investigations mentioned will be summarized and discussed. 2. NOMENCLATURE

The nomenclature introduced in the rst part of this paper must also be considered. - ai : minimum distance between zi and zi+1 axes; - D: distance between the bearings of the cross; - di : diameter of shaft i; - f : friction coefcient; - xi yi zi : moving cartesian system attached to the ith body, as in the Denavit-Hartenberg convention; - R(ik ) = CGi : vector expressed in C xik yik zik ; : Dual reaction forces acting on link j, RC j ( j) evaluated in C j , and expressed in the reference of the joint j. 1
( j)

- dual unity 2 = 0 - mechanical efciency of the Cardan jojnt; - i : angular velocity of the ith body, measured in the cartesian system o xiyi zi ; - 1 : Input angle; - f : frictional torque at the ith revolute joint; - the denote dual quantities; - Dots denote differentiation w.r.t. time; : Denote the skew matrix of a vector.
(i)

DYNAMIC ANALYSIS OF THE RCCC LINKAGE In terms of dual algebra the momentum of the ith rigid body w.r.t. a point C and expressed in the joint reference frame ik is HC(i)
(ik )

=mi vC(i)

(ik )

mi R(ik ) {i }(ik ) vC(i)

+ mi R(ik )

(i )

(1)

According to the Newton - Euler equation, differentiating Eqn.(1) w.r.t. time, one obtains the dual expression of the resultant of external loads (forces and moments) at point C acting on the rigid body d HC(i) dt
(ik )

= FC(i)

(ik )

(2)
(ik )

The joint reference frame ik is not an inertial frame and its motion is dened by means of the dual velocity vC(i) Thus Eqn.(2) is changed into HC(i)
(ik ) (ik )

+ vC(i)

(ik )

HC(i)

(ik )

= FC(i)

(ik )

(3)

is the time derivative of expression (1). where H C(i) Considering the RCCC linkage, the equilibrium equations of the moving links in terms of dual algebra are
2

A A A

1 3 2 4 3

RC1 (1) RC2 (2) RC3 (3)

(1) (2) (3)

RC2 (2) RC3 (3) RC4 (4)

(2) (3) (4)

= FC2(1) = FC3(2) = FC4(3)

(2) (3) (4)

, , .

(4a) (4b) (4c)

Using Eqns.(4) in Eqn.(3) one obtains the following system of linear equations (with j = 2, 3, 4) H C j ( j1)
( j)

+ vC j ( j1)

( j)

HC j ( j1)

( j)

= A

j j1

RC j1 ( j1)

( j1)

RC j ( j)

( j)

(5)

where the unknowns are the dual vectors of the reaction forces at the joint. Considering the system under investigation, the expressions of the dual momentum of each moving link can be obtained [4, 7]. 3. THE FRICTIONAL FORCES AT THE KINEMATIC PAIRS

In the ideal Cardan joint, the resultant of the reaction force at the kinematic pairs is zero and there is not any displacement along the joint axes. This is not necessarily true in presence of manufacturing errors. Considering that in a revolute joint the velocity along the z axis is zero (si = 0), the effects of friction on the reaction force component along the same axis (Fiz ) are herein neglected. However, these can be taken into account when detailed informations on the geometry of the revolute joint are available. 3.1 Revolute Pairs The frictional forces at the ith revolute joint arise from two sources: a) reaction forces Fxi , Fyi and Fzi ; b) reaction moments Mxi and Myi . The resistant action about the z axis, arising from the reaction moments Mxi and Myi , can be modeled [?, 9, 6] according to the scheme presented in Fig.1. In particular, for our purposes, the torque Mix is substituted by two parallel and opposite forces F acting normally to the revolute joint axis. Because of the presence of friction, these forces generate the frictional torque xi = f f di Mxi , Li (6)

where di is the diameter of the journal bearing, Li the distance between bearing supports1 and f the friction coefcient. yi d Similarly, the torque Myi generate the frictional torque f = f Lii Myi .
1 For

a single support bearing, Li represents the length of the bearing.

WITHOUT MANUFACTURING TOLERANCES

WITH MANUFACTURING TOLERANCES

20 0 -20 -40 -60 -80 -100 0 1 2 3 4 5 6

Mx1 My1 Mx2 My2 Mx3 My3 Mx4 My4 Mz4

50

-50

Mx1 My1 Mx2 My2 Mx3 My3 Mx4 My4 Mz4

Moments (Nm)

Moments (Nm)

-100

1 (rad)

1 (rad)

Figure 2: Comparison between torques at the kinematic pairs in joints without (left) and with (right) manufacturing tolerances Therefore the friction reaction torque component along the z axis is computed as follow Mzi = sign i
(i)

In the frictionless cylindric pair si = 0, Mzi = 0 and Fzi = 0. When friction is considered, the values of Mzi and Fzi must be computed.

Figure 1: Modeling of friction in revolute pairs

However, the computation of Mzi is carried out with the same model adopted for the revolute pair Mzi = sign i f +
(i)

f di 2

2 2 Fxi + Fyi

Fzi is dened in the same manner but making these substitutions Fzi = sign (si ) f

2 2 Fxi + Fyi + 2

2 2 Mxi + Myi

Li

The presence of friction in the kinematic pairs modies the equilibrium condition between links. In Fig.2 are compared the moments at the kinematic pairs with and without the mounting errors. The manufacturing tolerances are set to ai = 0.5 mm and i = 103 rad (i = 1, 2, 3, 4). 4. NUMERICAL EXAMPLES

The goal of this paper is to investigate the effects of manufacturing errors on the dynamic behavior of the Cardan joint. In particular the instantaneous efciency has been evaluated in several work conditions. The expression of the instantaneous efciency is Plost , (10) i = 1 Pin where Plost is the sum of power losses due to the frictional forces at each kinematic pair and Pin = 1 T1 is the input power due to the driving torque T1 . The power loss in each joint is dened as the product of the frictional force (9) and relative velocity s along the axis of the cylindric pair, and of the frictional torque (8) and the relative angular velocity . In the numerical examples herein reported the following nominal parameters have been assumed: - coefcient of dynamic friction f = 0.005; - lengths and diameters of the links Li =50 mm, di =40 mm, (i = 1, 2, 3); 3

Cylindric Pairs

d 2 2 where f = f Lii Mxi + Myi . It is clear that the presence of friction alters the equilibrium of the links.

f +

(i)

f di 2

2 2 Fxi + Fyi

(7)

(8)

(9)

1.000 0.975 0.950

=2 =4

=6
0.925 0.900 0.875 0.850 0.825

=8

=10

=1
=
=
24

20

=1

30

0.800 0.775 0.750 0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

0.22

0.24

a=2 f d/(L)

Figure 3: Average efciency in a Cardan joint without manufacturing tolerances under static conditions.[3, 5] - 1 = 2 = 3 = 90 , 4 = 150 and a j = 0, ( j = 1, 2, 3, 4); - masses of the links m1 =1.890 kg, m2 =1.57 kg, m3 =3.130 kg (the other inertial features are summarized in Tab.1); - input torque T1 = 100 Nm.

Table 1: INERTIAL FEATURES IN S.I. UNITS {S1 } {S2 }(3) {S3 }(4) [J1 ](2) x y z
(2)

x 0.117 0.081 0.127 x 0.01068 0 0

y 0.104 0.081 0.127 y 0 0.00784 0.00291

z -0.053 0 0 z 0 0.00291 0.00494

[J2 ](3) x y z [J3 ](4) x y z

x 0.05525 0 0 x 0.00928 0 0

y 0 0.005988 0 y 0 0.0111 0

z 0 0 0.001056 z 0 0 0.003826

The average efciency m of a Cardan joint, according to the model described in [6] can be obtained using the chart of Figure 3, where a = 2 f di /(Li ) is an adimensional parameter introduced by A. Morecki [3]. In the mentioned model, the Cardan joint has no manufacturing tolerances and the energy losses are computed including all the four revolute joints.

0,992 0,990 0,988 0,986 0,991

4= 165 4= 170 4= 175

0,984 0,982 0,980 0,978 0,976 0,974 0,972 0,970 0,968 0 1 2 3 4 5 6

0,990

i
0,989

1= 1500 rpm 1= 2000 rpm 1= 2500 rpm

0,988

0,987

0,986

0,985 0 1 2 3 4 5 6

1 (rad)

1 (rad)

Figure 4: Inuence of the angular speed on the efciency Figure 5: Inuence of the angular speed on the efciency and angular misalignement between input and output shafts The present analysis investigates the inuence on the efciency of the following parameters: 4

- Rotation speed. - Angular conguration of the output shaft. - Presence of manufacturing tolerances, estimated in the following values of axes offsets a1 = a2 = a3 = a4 = 0.5 mm and angular error on j =0.001 rad ( j = 1, 2, 3, 4). Figure 6 shows the comparison between instantaneous efciency loss (1 ) with and without manufacturing tolerances. The inuence on the mechanical efciency of rotation speed and of the angular displacement between input and output shafts has been also investigated. The results of this analysis are shown in Figure 4 and Figure 5. Finally the mechanical efciency of a Cardan joint with manufacturing tolerances or mounting errors has been computed. The angular velocity has been kept constant, 1 =1500 rpm, and the nominal parameters affected by errors have been changed into a1 = a2 = a3 = a4 = 0.5 mm and 1 = 2 = 3 = 90.006, 4 = 150.006. Figure 7 shows the results of this investigation. 5. EXPERIMENTAL ANALYSIS

The mechanical efciency model previously presented has been experimentally validated [10]. The test rig used, shown in Figure 8, was equipped with the following instruments: - an adjustable steel table; - two torque/speed transducers (model Magtrol TMB 210 with max torque: 100.00 Nm; max speed: 4000 r.p.m.; torque sensitivity 100mV/Nm; speed sensitivity: 60 pulses per rev.); - one brushless motor (two poles; peak torque: 110 Nm) with a control panel and control software; - one electromagnetic brake (model Merobel SA FRAT 650; max torque 65 Nm; min torque 0.63 Nm) with a radial fan and a DGT 200 MC digital controller; - one personal computer with an a/d converter and a National Instrument multichannel acquiring system.
0.070 With manufacturing tolerances No manufacturing tolerances

0,996

efficiency without errors efficiency with errors

0,995

Efficiency loss (1-)

0.065

Di
0,994 0,993

0.060

0.055

0,992
0.050

0,991
0.045 0 1 2 3 4 5 6

0,990 0 1 2 3 4 5 6

1 (rad)

G1 (rad)

Figure 6: Instantaneous efciency loss in a Cardan joint Figure 7: Inuence of manufacturing errors on mechanical efciency 6. CONCLUSIONS

Two formulations for computing the mechanical efciency of a Cardan joint with manufacturing tolerances or mounting errors have been presented. The rst formulation is based on quasi static working conditions. The results obtained coincide with those obtained by F. Duditza [3] for the standard Cardan joint, but with a different approach. Our methodology can take into account skewness between the axes of the joints. The second formulation, also based on dual numbers, includes the effects of inertia forces. Thus the effects on energy losses due to: - increment of the rotation speed; - angular errors between axes; - manufacturing tolerances; have been quantitatively analyzed. It can be observed that increasing the rotation speed from 1500 rpm to 2500 rpm the efciency loss is about 1.8%, while the efciency loss due to manufacturing tolerances is about 0.15%. This situation could be explained considering that the reaction forces at the kinematic pairs are scarcely affected by mounting errors, as shown in Fig.2. The model for computing the mechanical efciency under dynamic conditions have been also experimentally validated [10]. Also within the framework of dual numbers, paper [11, ?] discuss the effects of joint tolerances on the kinematics of spatial linkages. 5

Figure 8: Experimental test rig for the mechanical efciency analysis of mechanical joints References [1] Chen, C.K., Freudenstein, F., Dynamic Analysis of a Universal Joint with Manufacturing Tolerances, ASME Journal of Mechanisms, Transmissions and Automation in Design, vol.108, December 1986, pp.524-532. [2] Freudenstein, F., Macey, J.P., The Inertia Torques of the Hooke Joint, Proc. Of the 21st Biennial ASME Mechanisms Conference, Chicago, September 16-19, 1990, DE-Vol.24, pp.407-413. [3] Duditza, F., Transmissions par Cardan, Editions Eyrolles, Paris, 1971. [4] Fischer, I.S., Dual-Number Methods in Kinematics, Statics and Dynamics, CRC Press, Boca Raton, 1998. [5] Pennestr`, E., Vita, L., Mechanical Efciency of Cardan Joint with Manufacturing Tolerances, Proc. of the RAAD03 International Workshop, Cassino (Italy), 2003, Paper n.053RAAD03 [6] Biancolini, M.E, Brutti, C., Pennestr`, E., Valentini, P.P., Dynamic, Mechanical Efciency and Fatigue Analysis of the Double Cardan Homokinetic Joint, International Journal of Vehicle Design, vol.32, n.3/4, September 2003, pp.231-249 [7] Cavacece, M., Pennestr`, E., Valentini, P.P., Vita, L., Mechanical Efciency of a Cardan Joint, Proc. of ASME DETC04, Salt Lake City (USA), 2004, Paper DETC2004/MECH-57317 [8] Keler, M.L., Kinematics and Statics Including Friction in Single-Loop Mechanisms by Screw Calculus and Dual Vectors, ASME Journal of Engineering for Industry, May 1973, pp.471-480. [9] Dhanaraj, C., Sharan, A.M., Efcient Modeling of Rigid Link Body Dynamic Problems with Friction, Mechanism and Machine Theory, vol.30, 1995, pp.749-764. [10] Cavacece, M., Stefanelli, R., Valentini, P.P., Vita, L., A Multibody Dynamic Model of a Cardan Joint with Experimental Validation, MULTIBODY DYNAMICS 2005 Thematic Conference, Madrid, Spain, 21-24 June 2005 [11] Cecchini, E., Pennestr`, E., Stefanelli, R., Vita, L., A Dual Number Approach to the Kinematic Analysis of Spatial Linkages with Dimensional and Geometric Tolerances, Proc. of ASME DETC04, Salt Lake City (USA), 2004, Paper DETC2004/MECH-57324 [12] Pezzuti, E., Stefanelli, R., Valentini, P.P., Vita, L., Computer Aided Simulation and Testing of Spatial Linkages with Joint Mechanical Errors, International Journal For Numerical Methods in Engineering, vol. 65, pp. 17351748, 2006