Sunteți pe pagina 1din 29

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274317030

Kinematic Design and Multibody Analysis of Rzeppa Pilot-Lever Joint

Article  in  Proceedings of the Institution of Mechanical Engineers Part K Journal of Multi-body Dynamics · January 2008

CITATIONS READS

0 51

1 author:

Ettore Pennestri
University of Rome Tor Vergata
129 PUBLICATIONS   1,143 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Clifford algebra View project

Mechanism design View project

All content following this page was uploaded by Ettore Pennestri on 25 October 2015.

The user has requested enhancement of the downloaded file.


Kinematic Design and Multibody Analysis of
Rzeppa Pilot-Lever Joint

Ettore Pennestrı̀, P.P. Valentini

Dipartimento di Ingegneria Meccanica

Università di Roma Tor Vergata

00133 Roma - Italy

pennestri@mec.uniroma2.it valentini@ing.uniroma2.it

October 1, 2010

Keywords

Mechanical joints, Rzeppa joints, Kinematics, Multibody dynamics

Abstract

Rzeppa pilot-lever joint represents a classical solution to the problem of trans-


mitting torque between angularly misaligned shafts. Homokineticity is one

1
of the most important requirement in industrial applications. The Myard’s
theorem gives the geometrical conditions which ensure the satisfaction of
such requirement. In this paper two different original proofs of this theorem

are offered.
An algorithm for the optimal kinematic synthesis of the pilot-lever mech-
anism is then proposed. The pilot lever mechanism is designed to minimize
the maximum angle between the transmission and the homokinetic plane.

Finally, a multibody dynamic analysis of the Rzeppa pilot-lever joint is


made by means of a commercial software. The modeling of contact conditions
between the balls the guides is in particular discussed. Finally, the contact
forces between spheres and guides for different angles between input and
output shafts are computed.

1 NOMENCLATURE

• a1 and a2 : Axes of the shafts coupled with the joint;

• d: minimum distance between shaft axes;

• h and k: link length ratios;

• h∗ and k ∗ : optimal link length ratios;

• mball : mass of the ball;

• rin : inner shaft groove meridian curve radius;

2
• rout : outer shaft groove meridian curve radius;

• Sij : instantaneous screw axis of relative motion of body i with respect


to body j;

• Tout : Resisting torque on the driven shaft;

• α: Angle between the axes a1 and a2 ;

• αd : Angle between transmission and homokinetic planes or deviation


angle;

• ω1 and ω2 : Angular speeds of the coupled shafts;

• θ: Angle between the normal to the transmission plane and the a2 axis;

• θ1 and θ2 : Related angles of rotations of the coupled shafts;

2 DESCRIPTION OF THE RZEPPA JOINT

The Rzeppa joint type usage began in 1936 in front-wheel drive passenger

cars produced in the USA. Its distinctive capability is to transmit torque


between inclined shafts at almost constant velocity. However, it has been
also used in applications such as aircraft, marine, industrial stationary drive
systems.

3
a) b)

Figure 1: a) Pilot lever Rzeppa joint; b) Rzeppa joint

In Figure 1 two different types of Rzeppa joints, respectively patented in


1933 and 1934, are shown [1, 2].

In this paper, the so called pilot lever Rzeppa joint will be studied. The
joint connects two shafts B and D with intersection axes. In the following
the main components of the joint are briefly described.

• At the extremity of shaft B there is an outer hollow spherical member

4
named A in the drawing. This spherical member has a series of meridian
groove ball races whose circular axes are centered on the rectilinear axis
of shaft B.

• At the extremity of shaft D there is an inner spherical member named

C. This spherical member has also a series of meridian groove ball


races whose circular axes are centered on the rectilinear axis of shaft
D.

• Between the inner and outer spherical members there is a spherical


cage named E with circumferential slots receiveing the spheres which
engage the spherical slots of A and C.

• The angular position of E is adjusted by a pilot lever mechanism whose

purpose is to keep the plane containing the centers of the spheres (trans-
mission plane) as close as possible to the plane midway between the
plane of rotation of shafts B and D (homokinetic plane).

As will be demonstrated in this paper, the pilot lever mechanism cannot

ensure, for all relative angular positions of the input-output shafts, that
the all the centers of the spheres are contained in the homokinetic plane.
However, it will be possible to execute the kinematic synthesis of the pilot
lever mechanism in order to fulfill as close as possible the above geometric

condition.
In the pilot lever mechanism there is a force closure kinematic pair. In
fact, the axial spring, embedded in the axis B, maintains the contact between

5
kinematic elements.
In 1934 B.K. Stuber [3] devised a solution which avoided the use of the
pilot lever linkage. A.H. Rzeppa patented in 1939 an improvement of such

homokinetic joint [4].


The underlying principle of the solution proposed by Stuber is exemplified
in Figure 2.
Assuming no tolerances or manufacturing errors, the centers of the spheres

are kept always in the homokinetic plane. In fact, with reference to the ge-
ometry of Figure 2, points A and B are centers of spherical surfaces. Let
C be point of intersection of the shaft axes. This point maintains constant
and equal distances from A and B. Between the inner and outer spherical
surfaces there are two annular floating members. These bear upon opposite

sides of the balls which are in meridian grooves and transfer the torque from
one shaft to the other.
The radii of the spherical surfaces are such that, within the range of
feasible shafts angle, the distances of A and B from C are equal and constant.

Thus the condition of uniform transmission of rotation between shafts with


intersecting axes is guaranteed by the Euclid’s subtended angle theorem (see
right side of Figure 2).
It is interesting to observe that universal joints with pilot-lever linkages

have been also patented by other inventors, as reported in Table 1.


Classical monographies mainly dedicated to the kinematics and design of
the Rzeppa joint are authored by Wagner [5], Schmelz et al. [6], Duditza et

6
Outer frame

Inner frame

r2 r2 B
r1 r1

A C B A C

Figure 2: Scheme of the homokinetic joint patented by A.H. Rzeppa in 1939


and relative positions of the centers A and B of spherical surfaces

al. [7], Bongiovanni and Roccati [8].

3 THE MYARD’s THEOREM

The Myard’s theorem [9] states the condition required to obtain constant

transmission ratio in a mechanical joint. With reference to the geometry of

Table 1: Patents of universal joints with pilot lever linkages

Year US Patent Number Inventor

1935 2,010,809 A.H. Rzeppa


1940 2,211,388 D.A. Salvetti
1941 2,236,839 D.A. Salvetti
1967 3,357,209 J.J. Mooney Jr.

7
Figure 3, let the coupling transmit the motion between the shafts with axes
a1 and a2 . The axes of revolute pairs r1 and r2 always intersect in K. Since

C1 K = C1 C2 + C2 K (1)

and C1 K = h1 tan θ1 , C1 C2 = d and C2 K = h2 tan θ2 , then the relation

between the rotations θ1 and θ2 of the shafts is

d h1
tan θ2 = + tan θ1 . (2)
h2 h2

By time differentiation of (2), denoting with ω1 and ω2 the absolute an-


gular velocities of the two shafts, one obtains the transmission ratio

ω2 h1 1
τ= = " 2 # . (3)
ω1 h2 cos2 θ1 
d + h1 tan θ1
1+
h2

The homokinetic condition, (i.e. τ = 1), is always satisfied when d = 0 and


h1 = h2 . In other terms, it is required that:

• the axes a1 and a2 are incident (i.e. d = 0);

• point K is contained in the bisector plane (also named homokinetic


plane) of such axes.

Since the main elements of the coupling are spherically embodied, the first
requirement is matched by the Rzeppa pilot-lever joint. The purpose of the

8
K

A2

d
θ1 A1
D1
a2
r1 θ2
r2
a1 B1

C2 h2
h1

C1

Figure 3: Demonstration of the Myard’s theorem: Nomenclature

pilot-lever linkage is to maintain the transmission plane containing points

A1 , K and C1 as close as possible to the bisector plane. However, this last


requirement is not exactly fulfilled and for this reason the Rzeppa joint is
not homokinetic for any value of angle α.
With reference to the geometry of Figure 4, the deviation angle between
the homokinetic and transmission plane is

αd = |2θ − α| . (4)

9
α
θ
a2

π−α
2 π
a1 2

Homokinetic
Transmission plane
plane

Figure 4: Nomenclature

In the following sections it will be discussed a procedure for the optimal


kinematic design of this linkage. The objective of the kinematic synthesis is

to compute the proportions of the pilot lever linkage in order to keep αd as


small as possible for the feasible values of α. The closer is αd to the zero
value, the better the joint will approximate homokinetic working conditions.
By comparing the geometry of Figure 4 with the one of Figure 3 one can

conclude that

 
π−α  α
h1 = A1 C1 · sin − αd = A1 C1 · cos αd + , (5)
2 2
 
π−α  α
h2 = A1 C1 · sin + αd = A1 C1 · cos αd − . (6)
2 2

h1
Thus, the ratio h2
, which appears in (3) can be expressed as a function of αd

10
and α, is expressed by
cos αd + α2

h1
=  . (7)
h2 cos αd − α2

Substituting d = 0 and (7) into (3), one obtains a transmission ratio fully
consistent with the one deduced by I.S. Fischer et al. [10, 11], but through
a different procedure.

3.1 An alternative proof

A more concise proof of the Myard’s theorem can be obtained by considering

the spatial extension of the Aronhold-Kennedy theorem (e.g. [12, 13]).

y1 ≡ y3

1
2 S12

α
x1
z1
S13 d
3
θ

z3 S23

Figure 5: Spatial extension of the Aronhold-Kennedy theorem: Nomencla-


ture

With reference to the geometry of Figure 5, where bodies 1 and 2 are

11
connected to body 3 with cylindrical pairs, two body connected right hand
Cartesian references frames x1 y1 z1 and x3 y3 z3 are introduced. The axes z1
and z3 are directed along the axes of the cylindrical joints, whereas the co-

incident axes y1 and y3 are directed along the common perpendicular to z1


and z3 . Let θ and d be the angle and minimum distance between input and
output shaft axes, respectively. Distance d is measured along axes y1 and y3 .
The positions of the relative motions instantaneous screw axes are de-

noted with S13 (motion of body 1 with respect to body 3), S23 and S12 .
The space dual of the Aronhold-Kennedy theorem in plane motion states
that:

The axes of three relative motion instantaneous screws of three

bodies taken a pair at time interesect a common line normally.

Under homokinetic condition the relative angular velocities ω13 and ω23 of
body 1 and 2 with respect to body 3 (frame), are the same and the following
equalities must hold
ω13 = ω23 = ω .

The relative motion instantaneous screw axis S23 is in the intersection of


the following planes [12, 14]

sin θ z1
x1 = z1 = , (8)
1 − cos θ tan θ2
cos θ − 1
y1 = d . (9)
2 (1 + cos θ)

12
Since the transmission plane contains the common perpendicular to z1 and
z2 and S23 , the first equation clearly demonstrates that under homokinetic
conditions such a plane bisects the angle θ.

Other proofs of the Myard’s theorem, but substantially different from the
present one, are due to E. Metzner (1967) and M. Orain (1976) [6].

4 KINEMATIC ANALYSIS AND OPTIMAL

DESIGN OF THE PILOT-LEVER LINK-

AGE

4.1 Kinematic analysis

The actual kinematics of the pilot-lever Rzeppa joint is complex due to the
presence of multiple independent spatial closure loops. The kinematics is
further complicated because of tolerances and compliances.
For the purpose of kinematic analysis, in this investigation the Rzeppa

joint is simplified to one ball rolling inside two different toroidal grooves and
connected to a spherical cage.
The pilot-lever mechanism governing the angular position of the spherical
cage is composed of three spherical and two slider kinematic pairs.

In the plane of the two axes a1 and a2 , the pilot-lever linkage can be
interpreted as a planar six-bar function generator linkage [11] (see Figure
6a). In particular, assumed α as input variable, the output variable is θ.

13
The structural error of this linkage is αd as defined by equation (4).

a)

r2
r1
r3
α
θ β

r4
b)

Figure 6: a) Kinematic structure of the pilot-lever linkage; b) Main dimen-


sions nomenclature

From the geometry of Figure 6b) the following relationships can be es-

tablished:

sin β sin θ
= , (10)
r4 r3
sin (α + β) sin (α − θ)
= , (11)
r4 r23

where r23 = r2 − r3 . Substituting the first relationship into the second and

14
considering that s  2
r4
cos β = 1− sin θ , (12)
r3

after some manipulations, one obtains

2q sin 2θ − p cos 2θ + 2r + p = 0 , (13)

where
" 2  2 #  2
2 r 4 r4 r4 r4
p = tan α − − + , (14)
r23 r3 r23 r3
"  #
2
r4 r42
q = tan α + , (15)
r23 r3 r23
"  2 #
r4
r = tan2 α 1 − . (16)
r26

The solution of (13) is expressed by

p
−q + q 2 + r(p + r)
tan θ = (17)
(p + r)

15
4.2 Kinematic synthesis

The kinematic synthesis problem can be stated as follows: Compute the link
length ratios

r4
h= , (18)
r2
r3
k= , (19)
r2

such that the structural error αd is minimized within the feasible range of α.
In order to readily use (17), the following equalities are useful:

r4 h r4 1−k
= , = . (20)
r3 k r23 h

Table 2: Values of h∗ and k ∗ for αmin = 0◦ , αmax = 45◦ and maximum values
of the structural error.

h∗ k∗ αd max

0.7500 0.7085 2.14◦


0.6875 0.7265 1.99◦
0.6250 0.7453 1.85◦
0.5625 0.7649 1.72◦
0.5000 0.7856 1.60◦

16
The feasible ranges of h and k are, respectively,

0.50 ≤ h ≤ 0.75 , (21)

0.65 ≤ k ≤ 0.80 , (22)

whereas 0 ≤ α ≤ αmax .

Data: hmin , hmax , kmin, kmax , αmin , αmax


Result: Optimal values h∗ and k ∗
for h = hmin , hmax do

for k = kmin, kmax do


for α = αmin , αmax do
Compute θ by means (17);
Compute αd = |2θ − α|;

end

end

Obtain the values of h∗ and k ∗ which minimize the maximum value


of |αd | within the prescribed range of α ;

end
Algorithm 1: Kinematic synthesis of the pilot-lever linkage
By means of the synthesis procedure outlined in the Algorithm 1, different

optimal values of h∗ and k ∗ can be obtained (see Table 2). The Figure 4.2
shows the plot of αd .vs. α for h = 0.7500 k = 0.7085.

17
0.04

0.035

0.03
αd (rad)

0.025

0.02

0.015

0.01

0.005

0
0 0.1 0.2 0.3 0.4 0.5 0.6
α (rad)
Figure 7: Value of αd for h = 0.7500 k = 0.7085.

5 MULTIBODY ANALYSIS OF THE RZEPPA

JOINT

In order to simulate the dynamics of the Rzeppa ball joint, a three-dimensional


multibody model has been built using the CAD system Solidworks and its
Cosmos/Motion plug-in. The entire model is made of 11 rigid bodies: the

frame, two shafts, the cage, the pilot lever and six balls (see Figure 8). A
critical step of the modeling phase has been the simulation of the contact
between balls and the grooves of the two shafts. The implementation of
spatial contact between solid parts is computationally demanding. This is

18
Figure 8: The multibody model of the Rzeppa pilot lever joint

due to the presence of simultaneous impacts between balls and inner/outer

shafts which leads to excessive time for computation and discontinuous con-
tact forces. The analysis of this joint assuming only one ball in contact is
appropriate for an investigation limited to kinematics. The presence of n
balls will produce n − 1 dependent closure loops. In practice this redundancy

of constraints is avoided by introducing clearances between balls and grooves.


These functional gaps allow all the balls centers to be placed in the trans-
mission plane (the plane that contains all the ball centers) whose attitude is
governed by the pilot lever linkage, ensuring thus the mobility of the joint.
The equal spacing of balls favors load balance.

In order to avoid the redundancy of constraints in our multibody model,

19
Figure 9: Steps in building the virtual model

the interaction between balls and grooves has been simulated using force
elements instead of kinematic constrains. The balls have been rigidly con-
nected to the cage by means of spherical joints (see Figure 9a). Then, two

series of meridian curves have been sketched in the grooves (six curves on
the inner shaft and six curves on the outer shaft (see Figure 9b). Each pair
of curves intersects at a point which is contained in the transmission plane.
Two series of massless points have been created and constrained to these
curves (one for each curve). The point-on-a-curve constraint suppresses two

degrees-of-freedom. Finally, each ball has been connected with two spring-
damper elements to the corresponding points on the meridian curves (see

20
Ball 1 and 4 contact force with inner shaft

Ball 2 and 5 contact force with inner shaft


8.48
Ball 3 and 6 contact force with inner shaft

8.46

8.44

8.42
Force (N)

8.40

8.38

8.36

8.34

8.32

0 1 2 3 4 5 6

inner shaft rotation angle (rad)

Figure 10: Contact forces under homokinetic conditions (α = 20◦ )

Figure 9c).
The purpose of such elements is:

- to mimic the ball-groove contact elasticity;

- to model without overconstraining the multiple ball-groove contacts;

- to obtain an estimate of the actual contact load distribution.

This modelling approach does not take into account friction forces.

As concern the remaining kinematic constraints, both shafts are con-


nected to the frame with revolute joints. Moreover, the centers of the two

21
Test case No. α (deg) αd (deg) Model properties
0 20◦ 0.00◦ ω1 =2π rad/s
1 15◦ 1.75◦ Tout =2Nm
2 20◦ 2.07◦ rin = rout =40 mm
3 25◦ 2.10◦ mball =5 g

spherical ends of pilot lever are set coincident to the shaft axes (Figure 6a).
An input constant velocity (ω1 =60 r.p.m ) has been applied to the revolute
joint degree-of-freedom at the inner shaft and a resisting torque Tout = 2 Nm

has been applied to the outer shaft. Other inertial and geometrical param-
eters are reported in Table 5. The main purpose of the model is to assess
the contact forces between the balls and the grooves. Several scenarios have
been simulated by changing the relative angle between the two shafts and
they are herein discussed.

The first scenario is about the simulation of the transmission dynam-


ics when α = 20◦ and the structural error is 0◦ . In this configuration the
homokinetic plane (the bisector plane) coincides with the transmission plane
(i.e. the plane which contains the ball centers). This ideal scenario serves as

a reference for other simulations where αd 6= 0. Observing Figure 10 it can


be noticed that the contact forces between balls and grooves have a periodic
pattern even under homokinetic conditions. Their period is half the one of
inner shaft rotation. The angle phase-shift between two force plots is equal to

the angular spacing between the corresponding balls. The dynamic actions
on the grooves of inner and outer shafts are essentially the same.
Let us now discuss the simulation of the first optimal solution in Table 2

22
(h∗ = 0.7500; k ∗ = 0.7085), performed for the three configurations (test case
No. 0, 1 and 2) and summarized in Table 5.

Test Case No.1

Test Case No.2


1.016 Test Case No.3

1.012

1.008
Transmission ratio

1.004

1.000

0.996

0.992

0.988

0.984

0 1 2 3 4 5 6

inner shaft rotation angle (rad)

Figure 11: Transmission ratios computed with the multibody model

In Figure 11 the transmission ratios under dynamic conditions, for dif-

ferent test case analyzed, are reported. The plots agree with those obtained
from equation (3) when d = 0 and condition (7) is applied. Obviously the
transmission irregularities amplify when the structural error αd is increased.
They show a periodic pattern with a frequency which is twice the one of the

input shaft rotation.


As concern the interaction between the balls and the grooves, with ref-
erence to the Figure 12, the contact forces show a periodic pattern with the

23
Test Case 1 - Ball 1 and 4 contact forces with inner shaft

Test Case 2 - Ball 1 and 4 contact forces with inner shaft

Test Case 3 - Ball 1 and 4 contact forces with inner shaft


8.54

8.52

8.50

8.48

8.46

8.44
Force (N)

8.42

8.40

8.38

8.36

8.34

8.32

8.30

8.28

8.26

0 1 2 3 4 5 6 7

inner shaft rotation angle (rad)

Figure 12: Contact forces on inner shaft grooves for the test cases listed in
Table 3

same frequency of the transmission irregularities. Similar trend have been


observed for the remaining balls.

6 CONCLUSIONS

The paper discussed several technical issues associated with the dynamic

analysis and design of the pilot-lever Rzeppa joint.


A procedure of kinematic analysis of this joint and new proofs of the
Myard’s theorem have been proposed. One of these proofs was based on the
space extension of the Aronhold-Kennedy theorem. A kinematic synthesis

24
procedure for the design of the pilot-lever six bar linkage was suggested. The
design objective was the minimization of the angle αd between the plane
containing the ball centers (transmission plane) and the bisector plane of the

two shaft axes.


The multibody dynamic analysis of the joint allowed to estimate contact
forces between the balls and the grooves under dynamic conditions. Such
analysis is useful for the gathering of informations for strength and fatigue

design of all the joint components.


In particular, it was observed that the amplitude of contact forces depends
mainly on two parameters: the irregularities due to the difference between the
transmission plane and the homokinetic one, and the angular misalignment
between shafts. Maintaining the same structural error αd , the irregularities

grow by increasing shaft angular misalignement. When αd increases, the


reaction forces acting on the pilot lever become larger due to to the constraint
on the cage attitude. The contact forces between the balls and the moving
shaft are higher than those between the balls and the driven shaft. This

behaviour is due to the presence of torque oscillation at the input shaft,


caused by transmission irregularities, necessary to keep a constant input
velocity.
During the dynamic analysis of the Rzeppa pilot lever joint two disturbing

effects have been observed:

- a loading of elements with frequency double the one of shaft rotation;

25
- an increase with the structural error αd of the internal forces not di-
rectly contributing to torque transmission; (i.e. those acting on the
cage and on the pilot lever).

As a consequence of time-dependent loads, vibrations and fatigue phenomena


occur. The frequency of excitation is twice that of the input shaft rotation
and this has to be taken into account for a safe design and dimensioning.

References

[1] Rzeppa, A.H., Universal Joint, US Patent 2,010,899, 1933

[2] Rzeppa, A.H., Universal Joint, US Patent 2,046,584, 1934

[3] Stuber, B.K., Universal Joint, US Patent 1,975,758, 1933

[4] Rzeppa, A.H., Torque Transmitting Universal Torque, US Patent


2,150,942, 1939

[5] Wagner, E.R., Universal Joint and Driveshaft Design Manual, The So-
ciety of Automotive Engineers, 1991

[6] Schmelz, F., Seherr-Thoss, H., Aucktor, E., Universal Joints and Drive-
shafts, Springer Verlag, Berlin, 1991

[7] Duditza, F., D. Diaconescu, C. Jaliu, A. Barsan, and M. Neagoe. Cuplaje


Mobile Articulate. Editura Orientul Latin, Brasov (Romania), 2001

26
[8] Bongiovanni, G., Roccati, G., Giunti Articolati, Levrotto e Bella, Torino,
Italy, 1984

[9] Myard, F.E., Contribution à la géometrie des systémes articulés, Bul-

letin de la Société Mathématique de France, Vol. 59, pp. 183-210, 1931

[10] Fischer, I.S., Remington, P.M., Errors in Constant-Velocity Shaft Cou-


plings, ASME Journal of Mechanical Design, March 1994, vol.116,
pp.204-209.

[11] Fischer, I.S., Lawson, D., Kinematics of the Pilot-Lever Rzeppa


Coupling, Proceedings of the 4th National Applied Mechanisms and
Robotics Conference, Cincinnati, OH, Dec.1995, Paper AMR-102.

[12] Beggs,J.S., Advanced Mechanisms, The Macmillan Comapnay, New

York, 1966, p.62

[13] Bottema, O, Roth, B., Theoretical Kinematics, Dover Publications Inc.,


1990, p.106

[14] Cheli, F., Pennestrı̀, E., eds., Cinematica e Dinamica dei Sistemi Multi-

body, Casa Editrice Ambrosiana, Milano, 2006, p.74-79

List of Figures

1 a) Pilot lever Rzeppa joint; b) Rzeppa joint . . . . . . . . . . 4

27
2 Scheme of the homokinetic joint patented by A.H. Rzeppa in
1939 and relative positions of the centers A and B of spherical
surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Demonstration of the Myard’s theorem: Nomenclature . . . . 9


4 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
5 Spatial extension of the Aronhold-Kennedy theorem: Nomen-
clature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

6 a) Kinematic structure of the pilot-lever linkage; b) Main di-


mensions nomenclature . . . . . . . . . . . . . . . . . . . . . . 14
7 Value of αd for h = 0.7500 k = 0.7085. . . . . . . . . . . . . . 18
8 The multibody model of the Rzeppa pilot lever joint . . . . . 19
9 Steps in building the virtual model . . . . . . . . . . . . . . . 20

10 Contact forces under homokinetic conditions (α = 20◦ ) . . . . 21


11 Transmission ratios computed with the multibody model . . . 23
12 Contact forces on inner shaft grooves for the test cases listed
in Table 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

28

View publication stats

S-ar putea să vă placă și