Design, Analysis and Control of a

Spherical Continuously Variable Transmission

By
Jungyun Kim
Submitted in Partial Fulfillment of The
Requirements for The Degree of
Doctor of Philosophy
in the
School of Mechanical and Aerospace Engineering
at
Seoul National University
February 2001
To My Parents
i
Abstract
This dissertation is concerned with the design, analysis, and control of a novel con-
tinuously variable transmission, the spherical CVT (S-CVT). The S-CVT has a
simple kinematic structure, innitely continuously variable transmission character-
istics, and transmits power via dry rolling friction on the contact points between
a sphere and discs. The S-CVT is intended to overcome some of the limitations
of existing CVT designs. Its compact and simple design and relatively simple con-
trol make it particularly eective for mechanical systems in which excessively large
torques are not required.
We describe the operating principles behind the S-CVT, including a kinematic
and dynamic analysis. A prototype is constructed based on a set of design spec-
ications and results of theoretical performance analysis. In order to provide a
quantitative analysis of the spin loss of the S-CVT, which is one of the main sources
of power loss, we develop an explicit formulation using a modied classical friction
model, and an in-depth study of the velocity elds and normal pressure distribution
on the contact regions. The proposed friction model includes the pre-sliding eect,
i.e., Stribeck eects. Actual transmission ratios and power eciency are obtained
from experiments with a prototype testbench.
The open-loop shifting system of the S-CVT reveals nonlinearity and unstable
characteristics. In order to cancel the nonlinearity of the shifting system and to
make it stable to shifting commands, we develop an input-state feedback controller
based on exact feedback linearization. We also investigate the power eciency of
a generic dc motor, and present the results of a numerical investigation of the S-
CVTs energy savings possibility benchmarked against a standard reduction gear.
Furthermore, we develop a minimum energy control law for the S-CVT driven by a
ii
dc motor, and present numerical simulation results that conrm the performance of
the controller.
Finally, we design and construct an S-CVT based mobile robot to realize the
various advantages of the S-CVT into practical use. One of the key features of
the mobile robot is the design of a novel pivot mechanism for planar accessibility.
Results of both numerical simulations and experiments are presented to validate the
robots performance advantages obtained as a result of using the S-CVT.
Keywords: Continuously variable transmission; innitely variable transmission;
dry rolling friction; spin loss; feedback linearization; minimum energy control;
mobile robot.
iii
Contents
Dedication i
Abstract ii
List of Tables viii
List of Figures xi
1 Introduction 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 CVTs for Passenger Cars . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Outline and Contributions . . . . . . . . . . . . . . . . . . . . . . . . 19
2 Dynamic Analysis of the Spherical CVT 23
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.2 Kinematics of S-CVT . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.2 Operating Principles . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 Dynamics of S-CVT . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
iv
2.3.1 Motion of Sphere . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3.2 Shifting Dynamics . . . . . . . . . . . . . . . . . . . . . . . . 33
2.4 Reaction Forces of the S-CVT . . . . . . . . . . . . . . . . . . . . . . 35
2.4.1 Normal Reaction Force Exerted on the Variator: F
n
. . . . . 36
2.4.2 Shifting Reaction Force on the Sphere: D . . . . . . . . . . . 37
2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3 Prototype Design and Experimental Results 39
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2 Issues in Mechanical Design . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.1 Normal Force Loading Device . . . . . . . . . . . . . . . . . . 41
3.2.2 Capacity of Shifting Actuator . . . . . . . . . . . . . . . . . . 42
3.3 Prototype Specications . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.4 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.4.1 Performance of S-CVT . . . . . . . . . . . . . . . . . . . . . . 47
3.4.2 Strength and Life Prediction of S-CVT . . . . . . . . . . . . 49
3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4 Slip Analysis of the Spherical CVT 53
4.1 Friction Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Modied Friction Model for S-CVT . . . . . . . . . . . . . . . . . . . 59
4.3 Spin Loss of the S-CVT . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.1 Velocity Fields on the Contact Surface . . . . . . . . . . . . . 60
4.3.2 Normal Pressure Distribution . . . . . . . . . . . . . . . . . . 65
4.3.3 Quantitative Analysis of Spin Loss . . . . . . . . . . . . . . . 66
4.4 Slip Motion of the S-CVT . . . . . . . . . . . . . . . . . . . . . . . . 70
4.4.1 Stick and Slip States . . . . . . . . . . . . . . . . . . . . . . . 70
v
4.4.2 Slip Loss of the S-CVT . . . . . . . . . . . . . . . . . . . . . 70
4.4.3 Slip Involved Contact Analysis . . . . . . . . . . . . . . . . . 70
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5 Shifting Controller Design via Exact Feedback Linearization 74
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2 Stability Analysis of S-CVT Shifting System . . . . . . . . . . . . . . 76
5.3 Dierential Geometric Preliminaries . . . . . . . . . . . . . . . . . . 78
5.4 Shifting Controller Design via Input-State Linearization . . . . . . . 81
5.4.1 Controllability and Linearizability . . . . . . . . . . . . . . . 82
5.4.2 Input-State Linearization . . . . . . . . . . . . . . . . . . . . 83
5.5 Shifting Controller Design . . . . . . . . . . . . . . . . . . . . . . . . 84
5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6 Optimal Control of an S-CVT equipped Power Transmission 91
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.2 Power Eciency of a DC Motor . . . . . . . . . . . . . . . . . . . . . 93
6.2.1 DC Motor Dynamics . . . . . . . . . . . . . . . . . . . . . . . 93
6.2.2 Power Eciency of a DC Motor . . . . . . . . . . . . . . . . 95
6.3 Investigation of S-CVT Energy Savings . . . . . . . . . . . . . . . . 96
6.3.1 Control Design based on the Computed Torque Method . . . 98
6.3.2 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . 99
6.4 Minimum Energy Control via a B-Spline Parameterization . . . . . . 101
6.4.1 B-Spline Parameterization . . . . . . . . . . . . . . . . . . . . 102
6.4.2 Gradients of the Objective Function and Constraint . . . . . 103
6.4.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . 105
6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
vi
7 Case Study: An S-CVT based Mobile Robot 109
7.1 Motivation for Mobile Robot Applications . . . . . . . . . . . . . . . 110
7.2 MOSTS: An S-CVT Mobile Robot . . . . . . . . . . . . . . . . . . . 112
7.2.1 Pivot Device for Planar Accessibility . . . . . . . . . . . . . . 112
7.2.2 Prototype Design . . . . . . . . . . . . . . . . . . . . . . . . . 114
7.3 Numerical and Experimental Results . . . . . . . . . . . . . . . . . . 115
7.3.1 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . 118
7.3.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . 120
7.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
8 Conclusion 123
References 125
vii
List of Tables
3.1 Specications of prototype. . . . . . . . . . . . . . . . . . . . . . . . 46
3.2 Endurance test condition. . . . . . . . . . . . . . . . . . . . . . . . . 51
4.1 Maximal normal pressure comparison. . . . . . . . . . . . . . . . . . 66
5.1 Candidates for k
1
, k
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.1 Characteristic coecients of dc motor. . . . . . . . . . . . . . . . . . 97
6.2 Energy consumption; reduction gear vs. S-CVT. . . . . . . . . . . . 101
6.3 Energy consumption with the minimum energy control. . . . . . . . 108
7.1 Hardware specications of general mobile robots. . . . . . . . . . . . 110
7.2 DC motor charateristic coecients of MOSTS. . . . . . . . . . . . . 115
7.3 Energy consumption; MOSTS vs. dierential drive. . . . . . . . . . . 121
viii
List of Figures
1.1 Classication of transmissions for vehicles. . . . . . . . . . . . . . . . 2
1.2 Fuel consumption reduction for an engine. . . . . . . . . . . . . . . . 4
1.3 Engine speed variation. . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Classication of CVTs. . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Belts for belt drive CVT. . . . . . . . . . . . . . . . . . . . . . . . . 7
1.6 Belt drive CVTs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.7 Variable stroke drives. . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.8 Full toroidal CVT, by courtesy of Torotrak. . . . . . . . . . . . . . . 10
1.9 Structures for traction and friction drive CVT. . . . . . . . . . . . . 11
1.10 Geometries of toroidal CVT, by courtesy of Torotrak and NSK. . . . 13
1.11 Optimal operating line of an engine. . . . . . . . . . . . . . . . . . . 17
1.12 Typical variogram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.1 Standard structure of S-CVT. . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Velocity constraint diagram. . . . . . . . . . . . . . . . . . . . . . . . 27
2.3 Operating principles of S-CVT. . . . . . . . . . . . . . . . . . . . . . 28
2.4 Ideal speed ratio of S-CVT. . . . . . . . . . . . . . . . . . . . . . . . 29
2.5 Transmittable torque. . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.6 Coordinate system and forces on S-CVT. . . . . . . . . . . . . . . . 31
ix
2.7 Forces on variator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.1 3-dimensional concept view. . . . . . . . . . . . . . . . . . . . . . . . 40
3.2 Normal force loading device using a spring. . . . . . . . . . . . . . . 42
3.3 Schematic diagram of S-CVT. . . . . . . . . . . . . . . . . . . . . . . 44
3.4 Assembly drawing of S-CVT. . . . . . . . . . . . . . . . . . . . . . . 45
3.5 S-CVT prototype. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.6 Testbench of S-CVT. . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.7 Experimental results. . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8 Power eciency of S-CVT. . . . . . . . . . . . . . . . . . . . . . . . 49
3.9 Endurance test result of input disc. . . . . . . . . . . . . . . . . . . . 51
4.1 Spin loss in traction drives. . . . . . . . . . . . . . . . . . . . . . . . 54
4.2 Classical model of static, kinetic, and viscous friction. . . . . . . . . 55
4.3 Pre-sliding displacement phenomenon. . . . . . . . . . . . . . . . . . 57
4.4 Proposed friction model. . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.5 Contact of two bodies with dierent curvature. . . . . . . . . . . . . 61
4.6 Velocity vector eld on contact point. . . . . . . . . . . . . . . . . . 63
4.7 Typical relative velocity vector diagram. . . . . . . . . . . . . . . . . 64
4.8 Friction forces at the innitesimal area of the contact surface. . . . . 67
4.9 Spin losses on S-CVT at input speed of 3000 rpm. . . . . . . . . . . 69
4.10 Dislocation of contact center. . . . . . . . . . . . . . . . . . . . . . . 71
4.11 Change of normal pressure distribution in XZ plane. . . . . . . . . . 72
5.1 Stability of the S-CVT shifting system. . . . . . . . . . . . . . . . . . 87
5.2 Tracking performance of the S-CVT shifting system. . . . . . . . . . 88
5.3 Tracking error and corresponding control. . . . . . . . . . . . . . . . 88
x
5.4 System behaviors of S-CVT during the gear ratio change. . . . . . . 89
6.1 Diagram of an armature-controlled dc motor. . . . . . . . . . . . . . 93
6.2 Eciency of an armature-controlled dc motor. . . . . . . . . . . . . . 96
6.3 Target prole of output speed. . . . . . . . . . . . . . . . . . . . . . 97
6.4 Computed variator angle time prole. . . . . . . . . . . . . . . . . . 99
6.5 Motor behaviors; reduction gear vs. S-CVT. . . . . . . . . . . . . . . 100
6.6 Power consumption; reduction gear vs. S-CVT. . . . . . . . . . . . . 100
6.7 Interpretation of tilde07Eg(p). . . . . . . . . . . . . . . . . . . . . . 104
6.8 Optimal variator angle time prole. . . . . . . . . . . . . . . . . . . . 106
6.9 Motor behaviors with the minimum energy control. . . . . . . . . . . 107
6.10 Output behaviors with the minimum energy control. . . . . . . . . . 107
7.1 Pivot device for planar accessibility of MOSTS. . . . . . . . . . . . . 112
7.2 Electric circuit diagram of pivot switch and driving motor. . . . . . . 113
7.3 Hardware prototype of MOSTS. . . . . . . . . . . . . . . . . . . . . 116
7.4 The desired trajectory. . . . . . . . . . . . . . . . . . . . . . . . . . . 116
7.5 Calculated wheel velocity prole. . . . . . . . . . . . . . . . . . . . . 117
7.6 Trajectory of variator angle. . . . . . . . . . . . . . . . . . . . . . . . 118
7.7 Motor behaviors of MOSTS. . . . . . . . . . . . . . . . . . . . . . . . 119
7.8 Power consumption; MOSTS vs. dierential drive. . . . . . . . . . . 120
7.9 Experimental results. . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
xi
Chapter 1
Introduction
Power transmissions are a universal element in nearly all mechanical systems, from
a small-sized reduction gear in a compact disc drive, to a complex gear box (usually
referred to as a transmission) in a vehicle. Although their components, sizes, and
operating principles vary, their main objective is to eect changes in the sources
power in the manner that corresponds to the load condition by manipulating the
transmission ratio (or the gear ratio, i.e., the ratio of the input speed to output
speed). Well-designed power transmissions eliminate the need for oversized power
sources, and increase the power eciency of overall the system. Even though power
transmissions are required in various engineering elds, research activities are driven
mainly by automobile manufacturers for their conventional transmissions. Thus, in
this dissertation, an overview of power transmissions will be focused on automobile
applications.
1
1.1 Overview
The generated power from ordinary power sources (internal combustion engines,
electric motors, etc.) is much dierent from the necessary tractive force for driving
vehicles. Hence, it is necessary to transform the power adequately from the source
to the tractive force; the transmission of the vehicle takes this role. General trans-
missions for vehicles can be primarily classied into manual transmissions (MTs)
and automatic transmissions (ATs), according to its actuating mechanism for the
shifting action (decision of shifting time, engaging/disengaging of the power ow el-
ements, selecting the ratio, etc.). A detailed classication of transmissions is shown
in Figure 1.1.
A MT consists of dry clutch, which engages and/or disengages the power ow,
a pair of synchronizing devices and constant meshing gear train for each gear ratio,
and gear ratio selecting devices. Its structure and components are simple enough to
allow for a considerable reduction in size and weight compared to a conventional AT.
Transmission
Continuously
Variable
Transmission
Figure 1.1: Classication of transmissions for vehicles.
2
Moreover, a MT is built up with pure mechanical components and has no external
power loss, such as a hydraulic system; thus the power eciency is quite better than
that of an AT.
A conventional AT consists of wet clutches, planetary gear trains for each gear
ratio, a hydraulic system for shifting action, an electro-hydraulic servo system for
shifting control, and a torque converter. A torque converter is a unique device which
has the multiple roles of a torque multiplication device, starting device, and torsional
damper. Besides disadvantages in size and weight, a hydraulic system including a
torque converter shows signicant power loss, reducing the the overall eciency of
an AT (and ultimately the fuel economy of an AT equipped vehicle). However as
the driving comfort of vehicle becomes the main concern, and greater eort is made
toward improving the eciency of ATs, the market share of AT equipped vehicles
is growing rapidly.
Power sources have complex eciency characteristics according to driving con-
ditions. For example, an internal combustion engine has dierent fuel consumption
rates (or, brake specic fuel consumption: BSFC) according to its speed and torque
while producing the same amount of power (see Figure 1.2). In this gure, there
are two engine operating points which produce the same power for 120 km/hr with
regard to dierent gear ratios. In the case of gear ratio A, which is greater than B,
the BSFC value of this point is smaller than that of gear ratio B by 10%; hence one
can conclude that a wide-spread of gear ratios is helpful for improving a fuel econ-
omy. In addition, making more gear ratios can enhance the acceleration performance
for the same reason. Many transmission engineers therefore endeavor to develop a
transmission having more gear ratios. But making more gear ratios increases the
size and weight of a transmission.
The continuously variable transmission (CVT) has continued to be an object
3
E
n
g
i
n
e

t
o
r
q
u
e
Engine speed
100 %
110 %
120 %
130 %
140 %
150 %
Constant power
at 120 kph
Driving resistance
curve B
Driving resistance
curve A
BSFC curve
Engine torque
curve at W.O.T.
Vehicle speed 120 kph with gear ratio A
Vehicle speed 120 kph with gear ratio B
10% FC
reduction
Figure 1.2: Fuel consumption reduction for an engine.
of considerable research interest within the mechanical design community, driven
primarily by the automotive industrys demands for more energy ecient and en-
vironmentally friendlier vehicles. Unlike conventional stepped transmissions (MTs
and ATs), in which the gear ratio cannot be varied continuously, a CVT has a con-
tinuous range of gear ratios that can, up to device-dependent physical limits, be
selected independently of the transmitted torque. This feature of the CVT allows
for engine operation at the optimum fuel consumption point consistent with the
given output power requirements, thereby improving the engines power eciency.
Moreover, the CVT does not suer from shifting shock (see Figure 1.3).
In 1886, a CVT with rubber belt and pulleys made by Daimler Benz company
was known as the rst CVT to have been applied to a passenger car. About 1930,
General Motors acquired the patent of the toroidal drive, which will be mentioned
in a subsequent section in more detail, and tried to develop their own CVT. But
they failed to commercialize it, and nished the related research in 1935. A dierent
4
time
speed
time
speed
time
speed
Stepped
Transmission
CVT
Vehicle
Power source
repeat according
to the shifting
stay around
some point
regardless of shifting
Figure 1.3: Engine speed variation.
toroidal type CVT, known as a Heyes Self-Selector, was adopted in many Austin
cars, although its production ceased after two years.
The rst commercially successful CVT for a passenger car was the rubber belt
Variomatic of DAF Co., developed in 1958. The Variomatic was not popular, be-
cause it failed to resolve the problems of rubber belt failure and the performance
degradation due to deformation and wear. In the 1960s, a CVT using a metal belt
and variable pulleys was developed by Hub Van Doorne, but did not make it to the
market due to its insucient torque capacity.
In the 1970s, due to the worldwide oil-crisis and the raised environmental recogni-
tion, many countries strengthened the regulations of the fuel economy and exhausted
emissions of vehicles. Moreover by the advances of metallurgy and production tech-
nology, inherent restraints of CVT could overcome; the research and development
for CVT was much encouraged from the late of 1980s. Currently several automobile
manufacturers have developed various prototype CVTs that are soon expected to
5
appear in commercial vehicles (see [1]-[5] and references therein).
1.2 CVTs for Passenger Cars
According to the power transmission element and shifting mechanism, existing CVTs
can be classied into belt drive, traction drive, variable stroke drive, and hydro-
static/dynamic drive (see Figure 1.4).
Belt Drive CVT
In a belt drive CVT, a rubber or steel belt running on conically shaped variable
diameter pulleys is used to transmit power at dierent drive ratios. According to
the belt material, belt drive CVT can be divided into rubber, chain, and metal belt
type. In Figure 1.5, the schematic diagrams of rubber, chain, and metal belt are
shown. Because of their small power capacity, rubber belt CVTs are adopted in
Continuously
Variable Transmission
Friction Drive
Figure 1.4: Classication of CVTs.
6
(a) Rubber belt. (b) Chain belt. (c) Metal belt.
Figure 1.5: Belts for belt drive CVT.
compact cars and machine tools. Passenger cars equipped with a chain belt CVT
had previously appeared on the market, but their production halted before long
(a) ACVT of Aichi Co. (b) Multimatic of Honda Co.
Figure 1.6: Belt drive CVTs.
7
owing to chain noise and vibration problems.
Currently almost conventional CVTs have push type metal belts of Van Doornes
Transmissie b.v. or a revised form. Although the metal belt still suers from a
small torque capacity, the number of production units has been rising steadily in
the worldwide market. The torque capacity has recently increased with the aid of
advances in metallurgy and improvements in the hydraulic system. Figure 1.6 shows
the rubber belt drive CVT made by Aichi Co. and metal belt drive CVT by Honda
Co.
Hydrostatic/Dynamic and Variable Stroke Drive CVT
Hydrostatic/dynamic drives use an incompressible uid as the transmission medium,
by connecting a hydraulic pump directly to a variable displacement hydraulic actu-
ator. It can realize neutral, forward, and reverse stages, but is not typically applied
to passenger cars, owing to its own low power eciency, weight, and noise. It has
(a) Cylinder type. (b) Link type.
Figure 1.7: Variable stroke drives.
8
seen limited applications to heavy equipment.
A variable stroke drive is made with one-way clutches and crank devices which
can adjust the crank arm length. The rotational motion of the drive shaft transforms
into translational motion, and the one-way clutch recties the motion into a uni-
directional motion. This type of CVT cannot manage properly the pulsative output
torques, and is therefore not adopted in vehicles (see Figure 1.7, by courtesy of DOE
report [6]).
Traction and Friction Drive CVT
Friction wheels of unequal diameter were one of the earliest speed changing mech-
anisms. It is speculated that their use even predates that of gearing toothed
wheels, whose beginnings date back to the time of Archimedes, circa 250 B.C. [7].
Even today, friction drives may be found in equipment where a simple and eco-
nomical solution to speed regulation is required: phonograph drives, self-propelled
lawnmowers, or even amusement park rides driven by a rubber tire are a few of the
more common examples. In these examples, simple dry contact is involved, and the
transmitted power levels are low. However, this same principle can be harnessed
in the construction of an oil-lubricated, all steel component transmission which can
carry hundreds of horsepower using todays technology. In fact, oil-lubricated trac-
tion drives have been in industrial service as speed regulators for more than 70 years
[8].
Great progress in tribology research since late 1960s, particularly research on
elasto-hydrodynamic lubrication (EHL) traction, has made it easier to understand
the traction drive mechanism. Traction drives transmit power through an increased
shear force, which results from elasto-hydraulic shear stress of the traction oil be-
tween two rotating solid bodies. A coecient of traction is typically 0.1, and macro-
9
scopic slip occurs at any time. Since there is no direct contact between the rotating
bodies, wear phenomenon does not occur. The drive ratio is varied by controlling
the eective onset radius of the contact point.
According to the geometries of rotating elements, there are various type of trac-
tion drive CVTs: nutating drive, half toroidal type, and full toroidal type. The half
toroidal CVT has semi-circular discs, while full toroidal CVTs have full-circular discs
as input/output rotating elements. According to the curvature radii of discs, they
have dierent attainable gear ratios, torque capacity, and spin loss. Figure 1.8 shows
the full toroidal drive and pertinent CVT made by Torotrak Ltd. in UK. Apart from
this, many automobile manufacturers have developed half toroidal CVTs with dif-
ferent torque capacities in Japan. Along with the traction oil developers (Santotrak,
Shell companies), they have presented various prototypes of traction drive CVTs in
the market [9].
Generally, a traction drive shows rapid shifting response compared to belt drives,
(a) Full toroidal traction drive. (b) Pertinent CVT.
Figure 1.8: Full toroidal CVT, by courtesy of Torotrak.
10
and can be adopted to medium-sized (even to large-sized) vehicles, because the
highly-pressurized traction oil endures more shear stress than belt drives. The draw-
backs of a traction drive are known to be as follows: the need for careful temperature
control, sealing and supply of traction oil, and complicated shifting control due to
the three dimensional contact curvature of the rolling elements.
Finally, there exist friction drives where the power transmission mechanism is
via rolling resistance and friction force in direct contact, though its structure and
operating principle are much similar to traction drives (see Figure 1.9, by courtesy
of DOE report [5]). Friction drives have been also found in several types of wood-
working machinery dating back to before the 1870s. For example, [10] reports of a
frictional gearing being used to regulate the feed rate of wood on machines in which
one wheel was made of iron and the other, typically the driver, of wood (or iron
covered with wood).
Figure 1.9: Structures for traction and friction drive CVT.
11
Friction drives have not been considered for passenger cars due to its low torque
capacity, wear, heat dissipation problems, etc.However, friction drives have received
signicant attention from the perspective of tribology, because precise positioning
can be accomplished while avoiding backlash [11]-[15]. Furthermore, traction and
friction drives provide much design exibility in terms of their structure and al-
lowance for compact-sized designs.
Although each type of CVT has its own particular set of advantages and disad-
vantages, common diculties shared by current CVTs are the complicated shifting
controller design, and the need for a large-capacity, typically inecient shifting ac-
tuator [5]. Also, these CVT designs do not have innitely variable transmission
(IVT) capabilities, i.e., they do not include zero output speed among its available
ratios, and therefore require a clutch or other type of starting and engaging device
for initially driving the vehicle.
1.3 Literature Review
There is a vast amount of literature regarding the design, analysis, control, and
application of CVTs in engineering elds. This section focuses on the areas of
design and control of traction/friction drives, because the proposed spherical CVT
in this thesis shows similar characteristics with respect to operating principles, power
transmission and shifting mechanisms, and control laws.
Traction Drive Designs
It is well known that one of the earliest examples of the friction drive was the patent
of C. W. Hunt in 1877 [16]. Basically the mechanism of that drive was a toroidal
drive, which was developed for more than a decade thereafter. Applications of
12
traction drives to automobiles have been studied since the beginning of this century.
Prior to 1935, cars were called Friction Drive Cars, experimentally installed with
such drives, had received some attention: it was widely believed that power was
transmitted by friction between the rolling metallic elements.
In the latter half of the 1960s, when elsto-hydraulic lubrication (EHL) became
better understood [17], it was recognized that power was transmitted by traction.
The performance of a traction drive depends to a large extent on the rheological
properties of the uid in the EHL contact [18]-[29]. In the 1970s, synthetic traction
oil was developed which had a traction coecient almost 50% higher than before,
and practical use of the traction drive CVT was thought to be close at hand. It was,
however, not realized, because the heat treatment of the rolling elements could not
be achieved. A new type of traction oil being developed for automotive use shows
some promise [30], [31].
(a) Full toroidal CVT. (b) Half toroidal CVT.
Figure 1.10: Geometries of toroidal CVT, by courtesy of Torotrak and NSK.
13
As stated earlier, there are two main design streams in traction drives, for auto-
motive use full and half toroidal type CVTs. A full toroidal CVT [32] has full-circular
discs as input/output media and power rollers as shifting devices (see Figure 1.10
(a)). In a full toroidal CVT, power rollers are located at the center of the toroidal
shaped input and output discs. A hydraulic loading system has commonly been used
to supply the normal force, which is necessary to transmit power via traction. Its
shifting mechanism is based on the side-slip force generated by the velocity dierence
of the contact point.
A half toroidal CVT uses semi-circular discs instead full-circular ones, though
the shifting mechanism is not dierent from full toroidal CVTs (Figure 1.10 (b)).
Many engineers including P. W. R. Stubbs (1980), Lubomyr O. Hewko (1986), M.
Nakano (1991, 1999), H. Kumura (1999), and H. Machida (1999) have presented the
trends and issues on half toroidal CVT designs for use in full-sized cars as a future
driveline technology [33]-[38]. Nakano (1991) reported that the main reasons for the
unsuccessful commercialization of toroidal CVTs were thought to be the inability to
obtain sucient performance with respect to the traction and viscosity performance
of the traction uid, the fatigue strength of the rolling elements, power transmission
eciency, transient ratio change controllability, and the issue of synchronization
control in connection with the parallel arrangement of the traction elements [35].
A traction drive CVT changes its speed ratio by controlling the side-slip force
on the Hertzian contact area. Tanaka and Eguchi (1991) showed the principles of
the speed ratio changing mechanism of half-toroidal CVTs and highlighted a digital
compensation method for stabilization of the electro-hydraulically operated speed
ratio control mechanism [39]. Fellows and Greenwood (1991) reported that it might
not be possible to suppress hunting of the shifting control signal, depending on the
control system used [40]. In addition to these results, there are many materials
14
related to toroidal CVT controller designs (for example [41], [42], and references
therein).
When the ratio changes in a half toroidal traction CVT, the necessary contact
force does not vary signicantly compared to a full toroidal CVT [43]. Consequently,
a loading cam system of a half toroidal CVT that produces contact force in pro-
portion to the input torque can provide high eciency over the entire speed ratio
range, contrary to the hydraulic loading system of a full toroidal type. Moreover
it has been reported that the full toroidal traction CVT suers larger spin moment
at the contact points than the half toroidal type, which tends to reduce its power
capacity [44].
The current design issues on toroidal type traction drives can be summarized as
follows:
the material of rolling elements is not suciently reliable because of high pres-
sure and high temperature on the traction contact point;
there is no aordable traction oil which satises all the conditions of automo-
biles, although it has been reported that an adequate traction oil has been
developed recently [9], [31];
there are no bearings which can support high speeds and a large axial load;
the normal force loading system (e.g., hydraulics or loading cam), which is
necessary to produce traction force, is inecient and needs precise control for
equalizing the normal forces on the contact points;
there are diculties on the control of transient ratio change and synchronized
precise control in connection with the parallel arrangement of the traction
elements.
15
CVT Controls
A CVT is originally intended to operate the power source in power ecient regimes
by means of manipulating its gear ratio. Many previous eorts are focused on nding
the power ecient regimes of sources and controlling the gear ratio of a CVT in order
to run the source within those regimes. Here we review the previous analysis results,
which describe ways of controlling a CVT for maximizing the fuel economy of a
CVT-equipped vehicle as well as how to establish the shift schedule (the so called
variogram, which describes the graphic relation between the engine and vehicle
speeds) of a CVT for the vehicles performance objectives.
Generally the power eciency of a source is maximized at only one point over
its operating region. In an internal combustion engine (see Figure 1.2), the fuel
consumption is lowered for higher engine torque. On the other hand, it worsens for
high engine speeds as the mechanical loss is large at those speed points. The pumping
loss tends to be large for low engine speeds; hence, the fuel consumption also worsens
for low engine speeds. These characteristics are consistent with theory. If we operate
the engine only at the most ecient point, however, the driving performance may
not be satised, because the driving torque to be generated for each vehicle speed is
limited. Therefore, the control and optimization of automotive powertrain systems
with a CVT is achieved by cooperative control of the engine and CVT (see [2]-[4],
[45]). A drive-by-wire structure using an electric throttle control device is adopted
for this engine consolidated CVT control [46].
Figure 1.11 shows an optimal operating line (OOL) of a typical engine for max-
imum fuel economy. Generally, an OOL is constructed simply by connecting static
BSFC contours through optimization. For improving the fuel economy of a vehicle,
it is denitely helpful to control the CVT gear ratio so as to run the engine along
16
100 %
110 %
120 %
130 %
140 %
150 %
Figure 1.11: Optimal operating line of an engine.
this operating line. Most CVT-equipped vehicles use shift schedules (or variograms)
in look-up table form, presetting the optimal gear ratios obtained from the static
performance data of the engine and road tests of the prototype vehicle (see Figure
1.12). However, this OOL does not involve the vehicle dynamics including accel-
eration, performance objectives, because there is no consideration for the engine
dynamics.
The classical way to control CVT cars is to use some information on the gear
ratio or on the transmitted torque which is then fed back by a PID controller [47]-[49].
Only when using gain-scheduled controllers with typically 100 dierent gain points
could the required performance be achieved. Kolmanovsky et al. (1999) explored the
use of a CVT for torque management during mode transitions in lean burn gasoline
engines [50]. They demonstrated that an intuitively sound CVT gear ratio control
strategy which attempts to completely cancel the engine torque disturbance may
result in unstable zero dynamics. They concluded the coordination of engine torque
17
Figure 1.12: Typical variogram.
production and CVT gear ratio control during mode transitions was mandatory.
Takahashi (1998) proposed a scheme to minimize rate of fuel consumption by a
direct fuel injection engine used by combination with CVT [45]. Target values for
the engine and transmission which minimize fuel consumption ensuring driving per-
formance were calculated based on the nonlinear optimization method. As a result
of optimization, target values for air-fuel ratio and gear ratio were calculated and
controlled by tracking. Because the calculation of partial dierential was impossible
at some operating points, he used a simplex method that did not require calculat-
ing dierential values. For minimization of fuel consumption function under various
restrictions, penalty functions were also introduced.
The non-minimum phase behavior of the CVT based powertrain system (without
a torque converter) was mentioned in [51]. Considering this phase behavior of CVT,
Guzzella and Schmid (1995) addressed an exact feedback linearization approach for
a controller of CVT equipped vehicle [52]. In their works, the plant dynamics were
18
exactly linearized over the complete operating range using feedback linearization.
And as an application of the exact linearization approach, a kick-down controller
was designed.
1.4 Outline and Contributions
This dissertation deals with the design, analysis, and control of a spherical CVT. A
conceptual design of a particular spherical CVT (S-CVT) was proposed by Joukou
Mitsusida in [53]. The S-CVT consists of a sphere, input and output discs, and
variators. The rotating input and output discs are connected to the power source
and output shafts, respectively, while the sphere is situated between the input and
output discs. The transmission ratio is controlled by adjusting the location of the
variator on the sphere, which in turn controls the axis of rotation of the sphere. It
transmits power via dry rolling friction on the contact points of sphere and discs;
therefore, there exists a torque limitation decided by the static friction force.
The S-CVT, intended to overcome some of the aforementioned limitations of
existing CVT designs, is marked by its simple kinematic design and IVT charac-
teristics, i.e., the ability to transition smoothly between the forward, neutral, and
reverse states without the need for any brakes or clutches. Moreover its relatively
simple control makes it particularly eective for mechanical systems in which ex-
cessively large torques are not required (e.g., mobile robots, household appliances,
small-scale machining centers, etc.).
In order to put the S-CVT to practical use, an analysis of its operating principles,
power transmission and shifting mechanisms, and power capacity together with the
consideration for issues of hardware design, needs to be performed. This dissertation
is aimed at providing theoretical and practical solutions for these concerns, through
19
an in-depth study of the design, dynamics, and control of the S-CVT. This work can
be categorized into four parts:
analysis of theoperating principles, kinematics, and dynamics in Chapter 2;
hardware design issues including a slip loss analysis in Chapter 3, 4;
shifting controller and minimum energy control law design in Chapter 5, 6;
application for a wheeled mobile robot as a case study in Chapter 7.
The subsequent achievements in this work can be exploited to the design and analysis
of traction or friction drives having similar structure.
Currently, we are carrying out the development of other S-CVT applications
for small-capacity speed changers, e.g., bicycles, laundry machines, wind-propelled
generating systems, potters spinning wheels, etc.There still remain several practical
problems, such as realizing precise shaft alignments and increasing the torque capac-
ity. Currently research eorts are being directed toward the application of traction
uid for the purpose of adopting the S-CVT for large torque capacity applications,
e.g., hybrid vehicles, compact cars, etc.
The detailed outlines and contributions of each chapter can be stated as follows.
Operating Principles, Kinematics, and Dynamics
Chapter 2 describes the conceptual design and operating principles of the S-CVT
together with a detailed kinematic and dynamic analysis of its performance. In
addition, there shows analytic interpretation for the reaction forces of S-CVT which
are normally exerted on variator and discs, along with the denitions of their physical
meanings.
20
Hardware Design Issues and Slip Loss
Chapter 3 presents the prototype specications and a discussion of the main design
issues, focusing on the normal force loading device and the shifting actuator capacity.
Some experimental results are given on the actual transmission ratios and power
eciency obtained from a prototype testbench, to validate the operating principles
and performance of S-CVT. We briey address the strength and life estimation for
the S-CVT, based on the well-known ball-bearing life theory.
Spin loss of S-CVT, which is one of the main power losses of the S-CVT (and
more generally for friction and traction drives) due to slippage, is formulated using
a modied classical friction model in Chapter 4. The proposed friction model can
involve pre-sliding eect i.e., Stribeck eects. For this, we perform an in-depth
study of velocity elds and the normal pressure distribution generated on the contact
regions. We also provide a quantitative analysis of the spin loss of the S-CVT. In
addition, we discuss contact analysis involving slip, in which a shear force resulting
from friction occurs on the contact surface.
Shifting Controller Design and Minimum Energy Control
The shifting system of the S-CVT has second-order nonlinear dynamics, for which
typical open-loop control systems are likely to develop unstable characteristics. In
order to cancel the nonlinearity of the S-CVT shifting system and to make it stable
and responsive to shifting commands, we develop a feedback controller based on
the exact feedback linearization method in Chapter 5. We rst investigate the
instability of the S-CVT shifting system using the Lyapunovs indirect method. We
then present the input-state feedback controller design of the S-CVT shifting system,
and investigate the stabilizing and tracking performance of the dedicated shifting
21
controller by numerical simulation.
Chapter 6 deals with a minimum energy control law for the S-CVT connected
to a dc motor. We rst investigate the general power eciency of a dc motor. We
then present the results of a numerical investigation of the S-CVT energy saving
possibility benchmarked with a standard reduction gear. For this investigation, a
computed torque control algorithm for the S-CVT is proposed. In addition, we
describe a minimum energy control law of S-CVT connected to a dc motor. To do
this, we describe the general power eciency characteristics of a dc motor. Then
the minimum energy control design is carried out via B-spline parameterization.
Numerical results obtained from simulations illustrate the validity of our minimum
energy control design.
An S-CVT based Mobile Robot
Finally, we propose an S-CVT based mobile robot (denoted as MOSTS for a Mobile
rObot with a Spherical Transmission System) to realize the various advantages of
the S-CVT, including the originally intended CVT characteristic of energy eciency,
into practical use in Chapter 7. In this chapter, we rst address the motivation
for applying the S-CVT to a wheeled mobile robot by rst reviewing the current
hardware designs of mobile robots and their power eciency. We then present the
hardware design of our S-CVT based mobile robot in accordance with the target
performance. In addition, we propose a novel pivot mechanism which is necessary
for planar accessibility using an internal gear and an uncontrolled dc motor. We
perform both numerical simulations and experiments for various motion plans, in
order to validate the realization of the robots operation, the CVT characteristics,
and its energy saving possibility.
22
Chapter 2
Dynamic Analysis of the
Spherical CVT
2.1 Introduction
In this chapter a new type of spherical continuously variable transmission (S-CVT)
is described. The S-CVT, intended to overcome some of the aforementioned limi-
tations of existing CVT designs, is marked by its simple kinematic design and IVT
characteristics, i.e., the ability to transition smoothly between the forward, neutral,
and reverse states without the need for any brakes or clutches.
Because the S-CVT transmits power via rolling resistance between metal on
metal, it has limitations on the overall transmitted torque, which is eectively de-
termined by the static coecient of friction and the magnitude of the normal forces
applied to the sphere. Due to this torque limitation, the S-CVT is not intended
for automobiles and other large capacity power transmission applications. Target
applications for the S-CVT include mobile robots, household electric appliances,
23
small-scale machine tools, and other applications with moderate power transmis-
sion requirements. Although the current design of the S-CVT is based on friction
drive designs, it is our expectation that the power capacity of the S-CVT can be
increased by the use of traction oil, an issue which we do not pursue further in this
dissertation.
Other spherical CVT structures have been proposed for use in passive mobile
robots and for use as nonholonomic joints in robot manipulators. Carl A. Moore et
al. (1999) have reported a 3R passive robot, called the Cobot. The Cobot adopts a
rotational CVT to provide smooth, hard virtual surfaces for passive haptic devices in
place of conventional motors. Its rotational CVT consists of a sphere caged by four
rollers, and adopts the joint speeds and task space speeds along with the steering
angles as control inputs [54]. Another application can be found in underactuated
manipulators, designed by Serdalen et al. (1994). This work proposes a new type
of manipulator architecture using a CVT-type robot joint that takes advantage of
the inherent nonholonomy of the CVT [55]. Although these systems are designed
to manipulate the speed ratio using a CVT mechanism, their main purpose is not
for power transmission to improve the energy eciency. Furthermore, the shifting
mechanism of the S-CVT is quite dierent from these previous designs, as will be
described below.
In this chapter, the conceptual design and operating principles of the S-CVT
are described together with a detailed kinematic and dynamic analysis of its perfor-
mance. Section 2 describes the basic kinematic structure and operating principles
of the S-CVT. In Section 3, we examine the dynamics of the S-CVT by deriving
the equations of motion and its shifting mechanism. Finally in Section 4, we exam-
ine the reaction forces on the S-CVT, in particular those exerted normally on the
variator and discs.
24
2.2 Kinematics of S-CVT
2.2.1 Structure
The S-CVT is composed of three pairs of input and output discs, variators, and a
sphere (see Figure 2.1). The input discs are connected to the power source, e.g., an
engine or an electric motor, while the output discs are connected to the output
shafts. The sphere, which is the main component of the S-CVT, transmits power
from the input discs to the output discs via rolling resistance between the discs and
the sphere. The variators, which are connected to the shifting controller, are in
contact with the sphere like the discs, and constrain the direction of rotation of the
sphere to be tangent to the rotational axis of the variator.

Figure 2.1: Standard structure of S-CVT.
25
The speed and torque transmission ratios of the S-CVT vary with the angular
displacements of the variators; this will be described in further detail in the following
subsection on the operating principles of the S-CVT. To transmit power from the
discs to the sphere or from the sphere to the discs, a device that supplies a normal
force to the sphere, such as a spring or hydraulic actuator, must be installed on each
shaft. As can be seen in Figure 2.1, the structure and components of the S-CVT
are simple enough to allow for a considerable reduction in size and weight compared
to conventional transmissions. The orientations of the input and output shafts can
also be located freely using rollers at arbitrary positions rather than discs.
2.2.2 Operating Principles
When the input device is actuated by a power source, the input disc rotates about the
input shaft. This rotation in turn causes a rotation of the sphere, due to the condition
of rolling contact without slip between the input discs and the sphere. Rotation of
the sphere in turn causes a rotation of the output discs, and subsequently of the
output shaft. In the absence of any contact between the sphere and the variator, the
axis of rotation of the sphere will largely be determined by an equilibrium condition
among the various contact and load forces being applied to the sphere.
The role of the variator is to control the axis of rotation of the sphere. Specically,
referring to Figure 2.2, the variator contacts the sphere at a point (marked by P)
located directly above the sphere center. Since the variator rotates about an axis
normal to the variator disc and passing through the variator center (marked by C
1
,
C
2
), it follows that the contact point between the variator and the sphere undergoes
a linear velocity in a direction tangential to the variator disc (marked by V
1
, V
2
).
By adjusting the location of the variator (from C
1
to C
2
) it is therefore possible to
control the axis of rotation of the sphere (from
1
to
2
); the axis will be parallel
26
V
1
V
2
V

1

2
P
C
2
C
1
Figure 2.2: Velocity constraint diagram.
to the line between the variator center and the sphere-variator contact point, and
passing through the sphere center.
By varying the axis of rotation of the sphere, it is in turn possible to vary the
radius of rotation of the contact point between the input disc and the sphere, R
i
,
as well as the radius of rotation of the contact point between the output disc and
the sphere, R
o
(see Figure 2.3). In this way the speed-torque ratio of the S-CVT
can be adjusted. Figure 2.3 shows the various alignments of the variator for the
forward, neutral, and reverse states of the output shaft of the S-CVT. The neutral
state, which corresponds to zero rotation of the output disc, is achieved when R
o
becomes zero. As apparent from the gure, the forward, neutral, and reverse states
can all be achieved by smoothly manipulating the variator alignment, without the
need for any additional clutches or brakes.
Assuming roll contact without slip, the speed and torque ratio between the input
27
and output discs is related to the variator angle by the following relations:

out

in
=
r
i
r
o
tan (2.1)
T
out
T
in
=
r
o
r
i
cot (2.2)
where is the angular displacement of the variator,
in
and
out
are the respective
angular velocities of the input and output shafts, T
in
and T
out
are the respective
input and output torques, and r
i
and r
o
are the respective radii of the contact
points of the input and output discs (see Figure 2.3). There are two design variables
that prescribe the transmission ratio: the ratio of the input and output contact
r
i
r
o
R
R
i
R
o






r
i
r
o
R
R
i



r
i
r
o
R
R
i
R
o





variator angle
Figure 2.3: Operating principles of S-CVT.
28
-50
-30
-10
10
30
50
0.1
0.3
0.5
0.7
0.9
1.1
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Figure 2.4: Ideal speed ratio of S-CVT.
radii
r
i
r
o
, and the variator angle . From Equation (2.1) it is apparent that a large
range of available transmission ratios is possible even with a sphere of small radius.
Assuming that there is no slip or other physical eects, the ideal speed ratio of the
input speed to output speed is shown in Figure 2.4.
Although ideally an innite torque ratio is possible with the S-CVT as seen in
Equation (2.2), in practice there is a limit to the torque that can be transmitted
because power transmission occurs from rolling resistance of metal on metal. Figure
2.5 shows a plot of the torque ratio as a function of the variator angle, for a given
xed input torque. The actual torque ratio of the S-CVT will lie somewhere in the
operating region as indicated in the gure because of power loss due to friction, slip,
heat generation. The limiting torque T
max
is determined by the static coecient of
friction
s
and the normal force N exerted by the output disc spring mechanism on
29
the sphere according to the relation T
max
= r
s
N, where r is the contact radius of
the disc. When either the input or output torque applied at the disc-sphere contact
exceeds this limit, slippage can occur. Taking into account this limiting torque, the
output torque for a given input torque T
in
is given as follows:
T
out
= T
max
sat(
T
in
T
max
r
o
r
i
cot ) T
Loss
(2.3)
where the saturation function sat() is dened by
sat(x) =
_

_
sgn(x) if [x[ > 1
x if [x[ 1
,
and T
Loss
is the torque loss in S-CVT. Though assuming roll contact without slip
(i.e., the speed ratio can be realized as the ideal case), the torque loss cannot be
zero, because there exist some torque losses resulted from the spin moments and
internal loads, etc., which will be discussed in Chapter 3 and 4.
Figure 2.5: Transmittable torque.
30
2.3 Dynamics of S-CVT
2.3.1 Motion of Sphere
To investigate the shifting mechanism of the S-CVT, we designate a reference frame
XYZ situated at the center of the sphere, and a moving reference frame xyz, with
z coinciding with the spin axis of the sphere (see Figure 2.6). The various external
forces acting on the sphere are also shown in this gure, neglecting the normal
forces exerted on the sphere-discs contact points to hold the sphere and the weights
of sphere and discs.
We dene the driving forces which are delivered from the input discs as F
Zi1
and
F
Zi2
, and the reaction forces exerted by the load torque from the output discs as
Figure 2.6: Coordinate system and forces on S-CVT.
31
F
Zo1
and F
Zo2
. F
tv1
and F
tv2
denote the forces generated by the shifting actuator
acting at the sphere-variator contact points. The remaining reaction forces at the
input and output discs and variators are respectively denoted by F
Xi1
, F
Xi2
, F
Y o1
,
F
Y o2
, F
nv1
, F
nv2
.
Assuming that the sphere center does not move and the rotational axis of the
sphere lies on the xy plane, the force equilibrium conditions for each coordinate are
as follows:
_

_
F
Xo1
F
Xo2
+ F
Xi1
F
Xi2
+ (F
nv1
F
nv2
) cos (F
tv1
F
tv2
) sin
F
Y i1
F
Y i2
+ F
Y o2
F
Y o1
+ (F
nv1
F
nv2
) sin + (F
tv1
F
tv2
) cos
F
Zo1
F
Zo2
+ F
Zi1
F
Zi2
+ F
Zv1
F
Zv2
_

_
= 0. (2.4)
With respect to the specied coordinate frames, we can derive the dynamic equations
relating the angular momentum change with the resultant moment acting on the
sphere, i.e.,
d
dt
H
o
=

M
o
where H
o
is the angular momentum and

M
o
is the resultant moment. The
angular momentum of the sphere is given by:
H
o
= I
s
=
2
5
m
s
R
2
1
where is the angular velocity of the sphere, I
s
is its mass moment of inertia, m
s
the mass, and R the radius. Expressing the angular momentum of the sphere in
terms of the moving coordinate frames, we obtain the derivatives of this momentum
and the resultant moments, leading to the following set of second-order dierential
equations:
I
s
_

_
=
_

_
(F
nv1
+ F
nv2
)R (F
Zi1
+ F
Zi2
)Rsin (F
Zo1
+ F
Zo2
)Rcos
(F
Y o1
+ F
Y o2
+ F
Xi1
+ F
Xi2
)R
(F
tv1
+ F
tv2
)R (F
Zo1
+ F
Zo2
)Rsin + (F
Zi1
+ F
Zi2
)Rcos
_

_
(2.5)
32
where is the spinning rate of the sphere.
Considering the attributes of the external forces, as stated earlier, we can restate
those forces in Equation (2.5) as follows:
F
Zi1
+ F
Zi2
= F
i
,
F
Zo1
+ F
Zo2
= F
o
,
F
tv1
+ F
tv2
= F
t
,
F
nv1
+ F
nv2
= F
n
.
In the above equations, F
n
should not be regarded as an active force for shifting,
but rather as a loss-like force acting to resist any variator displacements. Examining
the reaction forces at the input and output discs caused by changes in the sphere
axis of rotation, we can also conclude that the magnitudes of these forces must be
equal, otherwise the sphere will be distorted:
F
Xi1
= F
Xi2
= F
Y o1
= F
Y o2
= D. (2.6)
The relevant forces can therefore be summarized as follows:
F
i
= Driving force delivered from the input discs;
F
o
= Reaction force caused by the output discs connected to the load torque;
F
t
= Shifting force on sphere delivered from the variator in the tangential direction;
F
n
= Loss-like reaction force exerted normally on variator;
D = Reaction force on sphere generated by the shifting.
2.3.2 Shifting Dynamics
To establish the dynamic relations between the sphere and variator, we rst dene
the forces on the upper sphere-variator contact point and the connected shifting
33
Figure 2.7: Forces on variator.
actuator (see Figure 2.7). To permit spinning motion of the variator, bearings are
located on the connecting rod, which connects the shifting actuator and variator.
In this gure, denotes the angular displacement of the shifting actuator, which
consists of the same number of variators, m is the mass of the variator, and the
eccentric distance between the centers of the shifting actuator shaft and variator. In
addition, F
sv1
is the shifting force delivered by the shifting actuator, a is the linear
acceleration of the variator center, and
v1
the rotational speed of variator.
Using the velocity constraint on the sphere-variator contact point, one can obtain
the rotational speed of variator
v1

v1
=

+
R

(2.7)
34
where is the spinning rate of the sphere, and R the sphere radius as dened earlier.
Let the moment of inertia of the shifting actuator shaft and connecting rod be I
a
,
and that of the variator be I
v
. The shifting force delivered from the variator onto
the sphere in the tangential direction (F
tv1
) can be written as
I
v

v1
= F
tv1
. (2.8)
By the force relation F
sv1
= ma+F
tv1
, and using the fact that the linear acceleration
of the variator a =

, as well as Equations (2.7), (2.8), we can express the shifting
torque F
sv1
,
F
sv1
= (I
v
+ m
2
+ I
a
)

+ I
v
R

. (2.9)
We assume that the lower variator always runs synchronously with the upper
one; then the total shifting force F
s
= 2F
sv1
and
v1
=
v2
. Rearranging the
equations of the sphere and variator (2.5) and (2.9), we obtain the following set of
second-order dierential equations for the S-CVT:
_
_
2
(I
a
+I
v
+m
2
)

2
RI
v

2
2
I
v

I
s
R
+ 2
RI
v

2
_
_
_
_

_
_
=
_
_
F
s
F
i
cos F
o
sin
_
_
. (2.10)
The reaction forces are given by
F
n
=
I
s
R

+ F
i
sin + F
o
cos , (2.11)
D =
1
4
I
s
R

, (2.12)
F
t
= 2
I
v

+
R

). (2.13)
2.4 Reaction Forces of the S-CVT
The two main reaction forces of the S-CVT are F
n
and D, which are exerted respec-
tively at the contact points between the sphere and discs, and sphere and variators.
35
In order to prevent slippage, they must be smaller than the maximal friction force.
In this section we derive analytic expressions for F
n
and D, and examine their eect
on the performance of the S-CVT.
2.4.1 Normal Reaction Force Exerted on the Variator: F
n
F
n
, the reaction force which is exerted normally on the variators, can be considered
as a loss force, and restricts the available gear ratios. Since F
n
acts ultimately on
the bearings located within the connecting rod, which connects the shifting actuator
and variator (see Figure 2.7), it can therefore cause excessive bearing normal forces
and bending moments on the variator and connected shafts.
From Equation (2.11), F
n
at steady state becomes
F
n
= F
i
sin + F
o
cos . (2.14)
From the fact that the shifting eort F
s
is zero at steady state, the relation between
F
i
and F
o
of Equation (2.10) becomes
F
i
cos = F
o
sin .
Substituting F
i
into Equation (2.14), F
n
becomes
F
n
=
F
o
cos
. (2.15)
Beyond a certain variator angle, the magnitude of F
n
becomes larger than the max-
imal friction force which is determined by the static coecient of friction
s
and
the normal force N; slippage therefore occurs at the sphere-variator contact point
(similar to the limiting torque T
max
).
During transient states, the dynamics of sphere and variator

inuences the
magnitude of F
n
additionally. More than any other reaction forces on the S-CVT,
36
F
n
varies considerably together with input/output force and the shifting dynamics;
thus it contributes the limit of available gear ratios of S-CVT. The allowable range
of F
n
during transient states is
F
n
=
F
o
cos
+
I
s
R

s
N .
Rearranging this, we obtain a range for the gear ratio :
[ [ cos
1
_
F
o

s
N
I
s
R

_
. (2.16)
In order to increase the range of available gear ratios, one can reduce the internal
load and hence increase the output force F
o
, or decrease the shifting response

, as
well as improve material properties with respect to
s
, N.
2.4.2 Shifting Reaction Force on the Sphere: D
There are four contact points between the sphere and input/output discs in the S-
CVT (see Figure 2.1). When shifting (i.e., changes in gear ratio) occurs, the reaction
force D, which resists the angular momentum change of the sphere, appears at each
contact point. The reaction force D is normally exerted on the discs, and it acts
directly on the bearings located within the input/output shafts; thus it can be
considered as loss force like F
n
.
Moreover from Equation (2.12), D is related with the shifting response

, and acts
to restrict the available shifting response. As is the case for F
n
, slippage resulting
from a reaction force larger than the maximal friction force makes the S-CVT unable
to transmit power; therefore the following inequalities must be hold:
D
s
N ,


4R
I
s

s
N . (2.17)
From this relation, we can conclude that the shifting response is constrained by the
material properties
s
, the sphere geometries R, I
s
, and the normal force N.
37
2.5 Summary
In this chapter, we have presented the design and analysis of a newly developed
spherical continuously variable transmission (S-CVT) focusing on its basic structure
and operating principles, shifting mechanism, and its reaction forces. The S-CVT is
intended to overcome some of the limitations of existing CVTs, e.g., dicult shifting
controller design, and the necessity of a large-capacity and typically inecient shift-
ing actuator. It is marked by its simple conguration, innite variable transmission
(IVT) characteristics and realization of forward, neutral, and reverse states without
any brakes or clutches.
Because the S-CVT transmits power through rolling resistance between metal on
metal, torque limitations prevent current versions of the S-CVT from being applied
to large capacity power transmission systems like passenger cars. However, our
study suggests that it can be well-suited for applications involving small mechanical
systems such as mobile robots, household electric appliances, small-scale machining
centers, etc.
Finally, we have investigated the reaction forces which are exerted normally on
the variator and discs. Both F
n
and D constitute sources of power loss for the
S-CVT; in particular, the magnitude variation of F
n
along the variator angle is
steeper than any other forces on S-CVT. Moreover F
n
can be a dominant factor in
determining the available range of gear ratios of the S-CVT. The shifting reaction
force D is related with the shifting response

and acts to restrict the available
shifting response.
38
Chapter 3
Prototype Design and
Experimental Results
3.1 Introduction
In designing a transmission, one must consider both the power capacity of the trans-
mission and power source as well as the load conditions. In this chapter, we rst
dene the design objectives of the S-CVT, taking into account its inherent charac-
teristics such as the power transmission mechanism based on friction force, shifting
mechanism, and operating principles.
The proposed S-CVT is intended for use in small capacity mechanical systems,
e.g., mobile robots, household electric appliances, small-scale machine tools, and
other applications with moderate power transmission requirements. In determining
the hardware specications of the S-CVT, practical issues such as the amount of
normal force required to assure rolling resistance at the contact points of the S-
CVT, and the capacity of the shifting actuator that can realize the desired shifting
39
response, and the range of available gear ratios must all be considered. Based on the
kinematic and dynamic analysis results of the previous chapter, we have designed
and built the following S-CVT prototype.
In this chapter, we present the prototype hardware specications for the S-CVT,
and an analysis of its performance. Using a prototype testbench, we obtain ex-
perimental results that serve to validate the operating principles and performance
of the S-CVT. In Section 2, we discuss various issues in the mechanical design of
the S-CVT, focusing on the normal force loading device and the shifting actuator
capacity. To assure rolling resistance force at the contact points of the S-CVT, we
adopt compressible springs because of their simple structure and the ease in ad-
justing the preset load. To determine the shifting actuator capacity, we derive the
Figure 3.1: 3-dimensional concept view.
40
numerical relationship between the necessary power and the shifting demand using
the previous dynamic analysis results. Section 3 shows the hardware specications
and schematic drawings of the prototype S-CVT. In Section 4, we present experi-
mental results on the actual transmission ratios and power eciency obtained from
the prototype testbench. Finally, we briey address the strength and life estimation
of the S-CVT, based on the well-known ball-bearing life theory.
3.2 Issues in Mechanical Design
Among the relevant issues in designing the S-CVT, we will focusing in particular
on the normal force loading device and the capacity of the shifting actuator. The
considered issues are mainly related to power capacity, namely the maximal trans-
mittable force and the shifting actuator design.
3.2.1 Normal Force Loading Device
In order to assure rolling resistant force at the contact points of the S-CVT, an appro-
priate normal force should be applied on the sphere and discs. Compressible springs
are employed at each shaft, which are connected to the variators and input/output
discs, to make the mechanical structure simple and to adjust the amount of normal
force easily (see Figure 3.2). Since the spring force is closely related with the limit
of transmittable force, we need to measure and adjust it. Using a set-screw, the
amount of normal force can be adjusted by xing the preset displacement of the
spring. To set an accurate spring force, strain gauges are attached to each relevant
shaft.
Regarding the amount of normal force, the larger spring force increases the trans-
mittable force. However, applying too large normal force causes yielding and plastic
41
Figure 3.2: Normal force loading device using a spring.
deformation of the sphere and discs; careful consideration for the normal stress on
the contact regions must be carried out. In this study, we have designated the nor-
mal force amount as 100 kgf using the corresponding nite element analysis results
obtained by ANSYS.
3.2.2 Capacity of Shifting Actuator
In order to determine the capacity of the shifting actuator, it is necessary to inves-
tigate the variation of shifting force F
s
along with the desired performance. From
Equation (2.10), in steady state F
s
is zero and the input-output force relation be-
comes F
i
cos = F
o
sin . To achieve shifting (i.e., gear ratio change), a non-zero F
s
must be induced by the shifting actuator in an appropriate manner.
For example, we consider the case when shifting occurs by the amount

d
at a
42
certain steady state instant. At the beginning of shifting, we can assume that the
input-output force relation still holds. Rearranging Equation (2.10), the shifting
force F
s
becomes
F
s
= 2
_
(I
a
+ I
v
+ m
2
)

2 I
2
v
R
2
(I
s

2
+ 2 I
v
R
2
)
_

d
. (3.1)
As seen in Equation (3.1), F
s
necessary for shifting is determined by R,

d
, and
the mass moments of inertia of the sphere I
s
, variator I
v
, and connected elements
I
a
+ m
2
. Considering that shifting forces of other traction or belt drives must be
large enough to resist the traction or friction force, which is generated directly by
the transmitted torque, the overall magnitude of shifting force of the S-CVT will
likely be much smaller than that of other existing CVTs.
The necessary power P
s
of the shifting actuator is calculated using Equation
(3.1):
P
s
= F
s

d
= 2
_
(I
a
+ I
v
+ m
2
)
2I
2
v
R
2
(I
s

2
+ 2 I
v
R
2
)
_

d
(3.2)
where

d
is the corresponding angular velocity to the required shifting demand

d
.
3.3 Prototype Specications
Based on the numerical investigation results from the previous studies, we have des-
ignated the hardware specications of the S-CVT prototype. The overall layout of
the power transmission is shown in Figure 3.3. Because of the maximum limiting
torque, a reduction gear with a ratio of three is added to the prototype; this ratio
also includes a safety factor. This additional reduction gear can be eliminated by
improving the material properties such as the static coecient of friction and in-
creasing the normal force at the contact point. The nal assembly drawing is shown
43
Driving Motor
Pivot
Motor
z=36
z=18
z=18
z=18
z=36
z
=
7
2
z
=
3
6
z
=
2
6
z
=
5
2 z
=
2
6
z
=
5
2
z
=
2
6
z
=
3
9 z
=
2
6
z
=
3
9
Variator
Sphere
Wheel
Wheel
10
Variator
Motor
z=18 z=66
z=66 z=18
1
3
13
Input disc
Figure 3.3: Schematic diagram of S-CVT.
in Figure 3.4. In Figure 3.3 and 3.4, a dc motor referred to as the pivot motor, and
internal gears are included in the power-ow line of the S-CVT. These elements are
added in order to make each output shaft rotate in opposite directions. This novel
pivot mechanism is proposed for the application to the CVT-based mobile robot,
which is the main subject of Chapter 7.
44
load spring
housing
Figure 3.4: Assembly drawing of S-CVT.
45
Element Nomenclature and Specication Material
Sphere
mass (m
s
) = 0.882 kg
radius (R) = 30 mm
Steel ball
of ball bearing
Input/output disc
Variator
mass (m) = 0.095 kg
radius (r) = 16 mm
contact radius (r
i/o
, ) = 10 mm
SNCM 8 class
Input/output shafts SCM 4 class
Gears Refer to Figure 3.3 SCM 21 class
Mass moments
of inertia
Sphere (I
s
) = 3.1758 10
4
kg m
2
Input parts (I
in
) = 2.3581 10
5
kg m
2
Output parts (I
out
) = 3.8609 10
4
kg m
2
Variator (I
v
) = 1.0514 10
5
kg m
2
Variator connected parts (I
a
) = 1.0585 10
4
kg m
2
Table 3.1: Specications of prototype.
The detailed specications for numerical studies and experiments are shown in
Table 3.1. The prototype S-CVT has been built and is shown in Figure 3.5.
Figure 3.5: S-CVT prototype.
46
3.4 Experimental Results
In order to validate the operating principles and performance of the S-CVT, we have
built a testbench for it. Two eddy-current type AC servo motors (input: 3-phase
AC, 122 V , 9 A; output: 1500 Watts; rated speed: 2000 rpm) are used for a driving
power source and a driven load generator. In the testbench (see Figure 3.6), the
variator angle is controlled by a dc stepped motor with an angular resolution of
0.024

/pulse. The rotational speeds of the input and output shafts are measured
through incremental optical encoders attached to the shafts.
3.4.1 Performance of S-CVT
Setting the external load torque to zero, we observe the output speed together with
the variator angle displacement while the input speed is set respectively to 748,
Figure 3.6: Testbench of S-CVT.
47
1502, and 2001 rpm. A steady-state speed ratio curve of the S-CVT is extracted
for the no-load condition (see Figure 3.7 (a)). Note that the overdrive of the output
speed, which implies that the output speed is faster than the input speed, occurs
when the variator angle exceeds 50

. In addition, there is a large deviation between

the ideal value and the test result beyond a variator angle of 65

, which indicates
the onset of slippage. These less than ideal output speeds arise from the increase of
reaction force normally exerted on variator F
n
, which is described in Section 2.4. In
the experimental result, moreover, there must be a certain amount of internal load
induced by manufacturing and other errors, which makes F
o
large (see more details
in Section 2.4).
Using slip-ring type torque sensors, we have also observed the output torque
together with the variator angle displacement by adjusting input/output torque to
realize the pre-obtained steady state speed ratio (see Figure 3.7 (b)). The actual
torque ratio is limited to under 20, which is determined mainly by the static coe-
0 10 20 30 40 50 60 70
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
(a) Speed ratio of S-CVT.
0 10 20 30 40 50 60 70
-5
0
5
10
15
20
25
30
35
40
45
(b) Torque ratio of S-CVT.
Figure 3.7: Experimental results.
48
0 10 20 30 40 50 60
0
10
20
30
40
50
60
70
80
90
100
Figure 3.8: Power eciency of S-CVT.
cient of friction and the exerted normal force.
Finally, we calculate the power eciency of the S-CVT using the obtained speed
and torque ratios (see Figure 3.8). The eciency is almost 85% for variator an-
gles under 15

, while the average eciency beyond this angle is about 65%. The
power eciency of the prototype S-CVT is somewhat low; this is mainly due to the
manufacturing errors including bearing friction loss, gear backlash, etc. From exper-
iments with the prototype S-CVT, we have also found that slight misalignments of
the shafts may cause bending moments in the shafts and discs, resulting in increased
bearing friction loss and slippage, although for applications accurate shaft alignment
will have to be separately addressed.
3.4.2 Strength and Life Prediction of S-CVT
Perhaps the most common form of mechanical failure in friction and traction drives
is by wear. The laws governing the overall friction and wear between two surfaces
49
seem to depend primarily on the total force transmitted across the two surfaces
rather than on the local distributions of stress and strain, and one may assume that
the two bodies in contact are perfectly rigid. In studying the details of the actual
mechanism of wear and friction, however, one must take into account the extremely
small areas of actual load contact between two bodies and the elastic and plastic
deformations in these regions. Another factor which must be considered in studying
the detailed mechanism is the surface condition of the metal, since this condition
may be such that these local points of contact behave in a manner quite dierent
from that of the same material in bulk form. It has been generally accepted that
the addition of a reasonable tangential force to a rolling contact has no appreciable
eect on drive life. This is so only when spin is almost entirely absent.
Dawe and Lohr (1993) reported that application of a realistic tangential trac-
tion force at the contacts does not seem to cause dramatic reduction in life, and
circular contacts appear to oer the best eciency concerning durability [56]. The
well known basic formula for the fatigue life of a rolling bearing was presented by
Lundberg and Palmgren (1947) [57]. According to the theory, the fatigue life of
rolling elements was caused by the maximum shear stress in the sub-surface of the
rolling contacts.
Machida et al. (1991) reported that rolling fatigue life of a traction drive is
inverse proportional to the cube of a load using the ball-bearing theory [30]. This
implies that a higher value of the contact force cannot be used simply to transmit
higher torque. Machida and Tanaka (1991) also presented an in-depth study with
experiments of oil lm and surface damage in traction drives [58]. With another
traction drive mechanism, Coy et al. (1981) presented a contact fatigue life analysis
method for multiroller traction drives [59]. The method was based on the Lundberg-
Palmgren analysis, and also used life adjustment factors for materials, processing,
50
Input torque range 0.5 - 4 kgf cm
Variator angle change 0 60

Normal force range 100 - 200 kgf

Table 3.2: Endurance test condition.
lubrication, and eect of traction.
Adopting these previous results, the strength and life prediction of the S-CVT
can be performed based on ball-bearing theory. However, there must be a prior
investigation of the normal and shear stresses at the contact points; these will be
discussed in the following chapter, with a detailed analysis of spin moments as well
as several numerical results by ANSYS.
Figure 3.9 shows the input disc surface after 10
7
revolutions under the following
test conditions (see Table 3.2). In this gure, there appears a circular deection
track along the contact points with the sphere. The deection width reaches almost
2 mm, and the depth reaches about 0.005 mm. Considering that the ideal elastic
Figure 3.9: Endurance test result of input disc.
51
deformation is 2.4 10
3
mm, further described in Section 4.3, the resulting stress
distribution (shear and normal stresses) must be beyond the elastic range. Moreover
from the wide deection track, each shaft has not been aligned exactly and bending
moment and slip loss must be resulted in the shafts and discs. However there is no
evidence of surface aking on the tested input disc.
3.5 Summary
In this chapter, we have discussed typical issues on the S-CVT hardware design,
focusing on the normal force loading device and the shifting actuator capacity. To
assure rolling resistant force at the contact points of the S-CVT, we adopted com-
pressible springs in the normal force loading device, because of their simple structure
and the ease in adjusting the preset load. To determine the shifting actuator ca-
pacity, we presented a numerical relationship between the necessary power and the
shifting response demand in explicit form, using the previous dynamic analysis re-
sults.
We presented hardware specications and some drawings for the S-CVT pro-
totype. In the prototype, an additional reduction gear was added for increasing
the transmittable limiting torque and internal gears for the mobile robot applica-
tion. Experimental results on the actual transmission ratios and power eciency
obtained with the testbench of S-CVT were represented in order to validate the op-
erating principles and performance. Finally, we briey addressed the strength and
life estimation for the S-CVT based on methods for predicting ball-bearing life.
52
Chapter 4
Slip Analysis of the Spherical
CVT
Like other friction and traction drives, slippage takes place whenever the transmitted
force (or torque) exceeds the limiting value at the contact points of the S-CVT.
Slippage causes wear, heat, power loss, and even failure of the power transmission;
therefore it is a critical problem in the design and control of the S-CVT. Although
many eorts have been dedicated to compensate for slip in a variety of mechanical
systems or ages, an accurate analysis of the mechanism of slippage is still an open
topic of research.
There are two main sources of power loss in the S-CVT, excluding the losses due
to its mechanical structure, bearings and shifting actuator (particularly in traction
drives, shifting actuator is composed of hydraulics). One is slip loss on the contact
points between the sphere and discs resulting from slippage in the rolling directions.
Once the rolling directional slippage occurs at these contact points, the transmitted
power becomes dierent from the desired value; the power transmission can even
53
fail.
The other one is called spin loss, which is also one of the main design issues
in traction drives also (see Figure 4.1) [60]-[62]. Spin loss results from the elastic
contact deformation of rotating bodies that have dierent rotational velocities. To
reduce the spin loss in traction and friction drives, many designers have investigated
dierent approaches to optimal contact geometry design, normal load application,
and controller design.
In this chapter a modied classical friction model, which describes pre-sliding
displacement friction, is presented to model the spin loss in the S-CVT. We also
perform a quantitative analysis of spin loss in the S-CVT along with an in-depth
study of velocity elds and the normal pressure distribution generated on the contact
regions. Finally we study the contact analysis of slip, when a shear force resulting
from friction occurs on the contact surface.

Figure 4.1: Spin loss in traction drives.
54
4.1 Friction Model Review
The study of friction has a long history in the elds of mechanics, metallurgy, tribol-
ogy, and control. Using both theory and experimentation, researchers have devel-
oped several dierent models of the structure and dynamics of friction. Pure rolling
friction conditions occur when the contact between two surfaces is a point. However,
according to Rabinowicz (1965), the contact region between two surfaces is typically
of larger area than a point due to an elastic (and possibly plastic) deformation [63].
Classical friction, also referred to as the stick-slip model , is the earliest and most
widely used model of friction. The three components of classical friction kinetic
friction, viscous friction, and static friction are illustrated on the friction versus
velocity graph in Figure 4.2. Although kinetic friction simply provides a constant
retarding force on rubbing surfaces, it also introduces a discontinuity at zero velocity.
As a result, servomechanisms performing bi-directional tasks will be subject to the
discontinuity during every velocity reversal. The discontinuous behavior of kinetic
Figure 4.2: Classical model of static, kinetic, and viscous friction.
55
friction can be classied as a hard nonlinearity; it is well-known that a closed
loop system with a hard nonlinearity can produce a limit cycle, i.e., self-sustained
oscillation, that would lead to poor control accuracy.
Viscous friction results from the viscous behavior of a uid lubricant layer be-
tween two rubbing surfaces. As shown in Figure 4.2, viscous friction is represented
as a linear function of slip velocity. Static friction is the force required to initiate
motion from rest. Typically, the magnitude of static friction is greater than that of
kinetic friction, which can lead to intermittent motion known as stick-slip. Stick-
slip manifests itself as repeated sequences of sticking between two surfaces with static
friction, followed by sliding or slipping with kinetic friction. In the servomechanism
control problem, stick-slip can diminish control accuracy; stick-slip limit cycling can
be avoided if damping and stiness are suciently high, however. The equations for
the classical lumped friction force model F
f
are as follows:
F
f
=
_

_
F
k
sgn(V ) + V if V ,= 0
F
s
sgn(F) if V = 0
. (4.1)
Here F
s
and F
k
denote static and kinetic friction, respectively. Equation (4.1) shows
that F
f
depends on the slip velocity V and coecient of viscous friction .
Contrary to the predictions derived from the classical friction model, researchers
including Courtney-Pratt and Eisner (1957) and others have found experimentally
that small relative displacements between two bodies in contact occur when the
applied tangential force is less than the static friction [64]. Dahl (1977) provided a
model of this pre-sliding displacement phenomenon [65], known as the Dahl model,
that assumes friction force is a function of displacement x and time t such that
dF
f
(x, t)
dt
=
F
f
(x, t)
x
x +
F
f
(x, t)
t
, (4.2)
56
with F
f
(x, t)/t = 0, and
F
f
(x, t)
x
= [1
F
f
F
fc
sgn( x)[
i
. (4.3)
and F
fc
are as shown in Figure 4.3 (a), and i is an exponent that Dahl empirically
derived to be approximately 1.5; however, it cannot be applied when the velocity
x 0.
While the simple static plus kinetic friction model oers an intuitive explanation
for the possibility of stick-slip oscillations, it does not oer adequate justication
for the existence of these limit cycles in the wide range of conditions under which
they have been observed. However, several researchers have found a source for this
discrepancy in the Stribeck eect, and experimentally derived a model of friction
variation with velocity as depicted in Figure 4.3 (b). The implication of the Stribeck
eect for servomechanism dynamics includes an increased likelihood of stick-slip limit
cycling at low velocities. Among many empirical models derived for the friction
(a) Dahl eect. (b) Stribeck friction.
Figure 4.3: Pre-sliding displacement phenomenon.
57
incorporating the Stribeck eect, the following is the most popular:
F
f
(V ) = F
k
sgn(V ) + V + (F
s
F
k
)e
(V/V
str
)
2
sgn(V ) (4.4)
where V
str
is the critical Stribeck velocity.
Leonard and Krishnaprasad (1992) presented a comparative investigation of fric-
tion compensating control strategies designed to improve low-velocity position track-
ing performance in the presence of velocity reversals for servomechanisms [66]. In
their work, the various controller designs incorporate dierent friction models rang-
ing from classical friction and Stribeck friction to the Dahl friction model. They
have claimed the Dahl model proved to be signicant for the friction compensating
control problem with repeated zero-velocity crossings.
Hu (1994) used Karnopps friction model, where the stiction zone is broadened to
an interval around zero velocity, for the position control of a servo-system containing
friction [67]. Lee et al. (1999) presented a numerical study with an extended Dahls
friction model [68]. In this study, they performed a comparative numerical analysis
of the computational cost and modelling eorts between the classical stick-slip model
and an extended Dahls friction model for an automatic transmission system analysis.
Besides the above analyses of friction mechanisms, control engineers have used
open-loop smoothing techniques, such as dither and pulse-width modulation, to
compensate for friction in mechanical systems. However, these techniques have
disadvantages, e.g., dither can cause mechanical problems such as fatigue by exciting
vibrations in manipulators. As an alternative to these techniques, recent works have
brought to the forefront adaptive and estimation-based control techniques for the
compensation of friction in mechanical systems [69]-[72].
Lee and Tomizuka (1996) presented a controller structure for robust high-speed/
high-accuracy motion control systems [13]. In their control system, the friction com-
58
pensator is based on the experimental friction model and compensates for nonlinear
friction that is not modeled. Vedagarbha, Dawson, and Feemster (1999) have de-
signed an observer-based exact model knowledge position tracking controller for a
second-order mechanical system with nonlinear load dynamics and the nonlinear
dynamic friction model proposed by C. Canudas et al. [73] in their recent work [74].
In traction drive elds, the slip of traction oil within the contact region is an
important issue in determining torque capacity, power loss, the shifting mechanism,
and mechanical life. Although the fundamental mechanisms of traction diers from
that of friction, the kinetic characteristics are similar; therefore many engineers use
the terminology of friction to construct numerical models for traction drives [18],
[19], [21], [23], [24], [27], [29], [30], [75].
4.2 Modied Friction Model for S-CVT
There is spin loss in S-CVT like traction drive CVTs; this is described in the fol-
lowing section in detail, where the pre-sliding eect in the vicinity of zero relative
velocity is considered. Friction models based on Dahls show diculties in numerical
integration due to the high stiness and damping coecients; moreover they must be
obtained through a careful experimental analysis [68]. Thus, the classical stick-slip
friction model is adopted to the S-CVT system, because velocity reversal seldom oc-
curs in the S-CVT. We propose a modied classical friction model including Stribeck
eect like
F
f
=
_
(
s

k
) exp(
V
V
str
)
2
+
k
_
P sgn(V ) (4.5)
where
s
,
k
are static and kinetic coecients of friction, respectively. Here we
neglect viscous friction, as there is no lubricant layer in the S-CVT. For typical
59
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
-10
-5
0
5
10
Figure 4.4: Proposed friction model.
values of friction model parameters, the friction force versus slip velocity of the
proposed model is depicted in Figure 4.4.
4.3 Spin Loss of the S-CVT
4.3.1 Velocity Fields on the Contact Surface
The friction force on the contact surface is determined by the normal force and
friction coecients. Considering that the kinetic coecient of friction is related to
the relative velocity V between two rotating bodies, we rst investigate the relative
velocity eld on the contact surface. In this subsection, Hertzian results for elastic
deection are employed to construct the geometric parameters of the contact surface.
Hertzian theory deals with the general solution of the elastic contact problem of rigid
bodies [76], [77].
60
Contact of Rotating Bodies with Dierent Radii of Curvature
Consider two solid bodies in contact under a normal force P (Figure 4.5). In this g-
ure, R
x1
, R
y1
and R
x2
, R
y2
denote their radii of curvature respectively. The material
properties (e.g., Youngs modulus and Poissons ratio) of each body are E
1
,
1
and
E
2
,
2
respectively. We set the reference frame XYZ to be on the contact point,
and the local coordinate frame z on the deformed contact surface such that its
origin coincides with the contact center. According to Hertzian analysis [77], the
contact surface in this case produces an elliptic shape with principal axes a, b and
corresponding normal deections
x
,
y
; these can be approximated as
a = 1.109
3

P
E

R
y1
R
y2
R
y1
+ R
y2
, b = 1.109
3
_
P
E

R
x1
R
x2
R
x1
+ R
x2
, (4.6)

x
= 2.64
3

P
2
E
2

R
x1
+ R
x2
R
x1
R
x2
,
y
= 2.64
3

P
2
E
2

R
y1
+ R
y2
R
y1
R
y2
, (4.7)
Figure 4.5: Contact of two bodies with dierent curvature.
61
2
E
=
(1
1
2
)
E
1
+
(1
2
2
)
E
2
, (4.8)
where E is the equivalent Youngs modulus.
Suppose that body 1 and body 2 have angular velocities
1
and
2
, respectively.
Here we assume that the contact center does not dislocate from the original contact
center (i.e., roll without slip in the rolling direction). Then we can derive the
relative velocity eld V(, ) = V
1
(, ) V
2
(, ) in the contact surface with
pure rotational motion in R
3
, using the relevant curvature radii as follows:
V(, ) =
_
_
(R
y1

y
2
)
1Y
(R
y2

y
2
)
2Y
+ (
1Z

2Z
)
(R
x1

x
2
)
1X
(R
x2

x
2
)
2X
+ (
1Z

2Z
)
_
_
T
(4.9)
where
iX,Y,Z
is the rotational velocity component in the reference frame of each
body.
Contact of Disc and Sphere: S-CVT Case
Figure 4.6 shows the contact surface between the sphere and upper variator. In this
gure, we set a local coordinate frame y, in the directions of xyz as shown in
Figure 2.6, at the center of the contact surface S; the rolling direction is in the
direction. Taking into account that the two contact bodies in the S-CVT are disc
and sphere, one can let R
x1
= R
y1
= and R
x2
= R
y2
= R in Equations (4.6)-
(4.8). Furthermore, supposing there is no bending deformation of the variator along
the x, z axes, the normal deection and contact surface radius c can be calculated
as
c = 1.109
3
_
P
E
R , = 2.64
3
_
P
2
E
2

1
R
. (4.10)
To obtain each velocity eld, we rst recall that the sphere has a pure rotational
speed of in the z direction and the variator a rotational speed of
v
in the y
62
Figure 4.6: Velocity vector eld on contact point.
direction. The velocity eld of the sphere V
1
(, ) and that of variator V
2
(, ) on
the contact surface can be obtained as follows:
V
1
(, ) = [ (R ), 0 ] ,
V
2
(, ) = [ ( + )
v
,
v
] ,
(4.11)
where is the distance between the contact surface center and variator center. Con-
sequently, the relative velocity eld V
v
(, ) can be derived as
V
v
(, ) = [ (R ) ( + )
v
,
v
] . (4.12)
Similarly, the relative velocity elds at the other contact points (input and output
discs) can be obtained.
63
As shown in Equation (4.12), there is a relative velocity component of V
v
in the
direction, V

, on the contact surface. However V

(, ) does not accumulate

total relative velocity in the direction, because it is symmetric along the axis.
Therefore, V

(, ) contributes to spin along the direction normal to the plane.

In the case of V

(, ), the relative speed of ( +

v
) occurs in the rolling
direction (recall the rotational speed relation of R =
v
). Note that becomes
small enough to be neglected compared R (for example, = 2.4 10
3
mm, c =
0.59 mm, and R = 30 mm for the case of the S-CVT prototype); therefore V

(, )
can be approximated to be
v
. The contribution of V

(, ) is also a spin, similar

to V

(, ).
The vector diagram of relative velocity is obtained using typical values of ,
v
,
, R, P, and E that correspond to the S-CVT prototype specication (see Figure
-0.0006 -0.0004 -0.0002 0.0000 0.0002 0.0004 0.0006
-0.0006
-0.0004
-0.0002
0.0000
0.0002
0.0004
0.0006
Figure 4.7: Typical relative velocity vector diagram.
64
4.7). The spin velocity eld can be found straightforwardly, although there are no
excessive forces that cause slippage. From this result, we can be assured that there
must be spin in the contact surface around the origin of the local coordinate frame
of the contact point in the S-CVT regardless of the existence of shear force resulting
slippage.
4.3.2 Normal Pressure Distribution
Now we consider the normal pressure distribution on the contact patch to obtain the
friction force as well as to analyze the strength of the S-CVT. A Hertzian pressure
distribution develops in the circular shaped contact patch (with radius c) between
the sphere and disc. The pressure at each point in the contact surface is known to
be
p (, ) =
3
2
P
c
3
_
c
2

2
. (4.13)
The maximal normal pressure p
max
is located at the center of the contact surface;
at the boundaries, the normal pressure p becomes zero. The mean value of normal
pressure, p
mean
, equals the normal contact force P divided by the area of the contact
surface:
p
mean
=
P
c
2
=
1
1.109
2

3
_
PE
2
R
2
.
Using the relation p
max
= 1.5 p
mean
, the maximal normal pressure can be calculated
as
p
max
= 0.388
3
_
PE
2
R
2
. (4.14)
The maximal pressure p
max
should not exceed the yield strength of the sphere
and disc. For design purposes, the maximal normal pressure values for several com-
binations of contact bodies are illustrated in Table 4.1. On the same load condition,
65
Table 4.1: Maximal normal pressure comparison.
the maximal pressure values for the S-CVT (i.e., sphere and disc contact) reaches
almost 1.588 times that of the case that is inscribed in a circular body more than
twice the radius.
4.3.3 Quantitative Analysis of Spin Loss
Relative velocities resulting from the elastic contact of rotating bodies usually give
rise to friction mechanisms; in which case friction moments (spin loss) occur in the
66
contact region. Once there is an elastic or plastic deformation at the contact points
of bodies with dierent rotational velocities, there must occur a spin loss. Spin loss
is not caused by changes in the applied or exerted forces, but by the dierence in
rotating velocities and geometric properties of the contact bodies. Therefore spin
loss always exists in traction and friction drives; only the amount is dierent for
each drive [5], [19], [20], [23], [27], [32], [33], [35], [36], [61], [62], [75].
Consider the innitesimal area at the contact surface S, with the friction force
of the ith area in the rolling direction ( direction) denoted F
i
, and F
i
the force
in the direction as shown in Figure 4.8. The total friction forces F

and F

can be
obtained using the following equations;
F

=
_
c
c
_
c
c
F
i
(, ) dd , F

=
_
c
c
_
c
c
F
i
(, ) dd .
Recall that the normal pressure distribution has symmetries along the and axes,
and that V

in Equation (4.12) varies along the direction (neglecting ), and V

along the direction; there are no total relative velocities in the and directions.
Figure 4.8: Friction forces at the innitesimal area of the contact surface.
67
Therefore one can conclude that F

and F

become zero.
Using the proposed friction model in Equation (4.5), the spin moment T
spin
for
the variator can be calculated as
T
spin
=
_
c
c
_
c
c
(F
i
+ F
i
) d d (4.15)
where
F
i
= [(
s

k
) exp(
V

V
str
)
2
+
k
] p(, ) sgn(V

) ,
F
i
= [(
s

k
) exp(
V

V
str
)
2
+
k
] p(, ) sgn(V

) ,
V

=
v
,
V

=
v
.
Rearranging and integrating by parts, Equation (4.15) becomes
T
spin
=
3Pc
4
_

k
+ (
s

k
)
2
c
4
V
str
2

v
2
_
c
2

V
str
2

v
2
(1 exp(
c
v
V
str
)
2
)
__
(4.16)
where c is the radius of the contact surface, which can be calculated using Equation
(4.10).
To investigate the amount of spin loss at the contact points of the S-CVT, we cal-
culate the respective spin losses using Equation (4.16) for the input and output discs
and variators with typical values of
k
,
s
, V
str
, P. Figure 4.9 shows the numerical
results for spin loss at an input speed of 3000 rpm along the variator angle change,
and the speed changes in the discs and variator. Spin moments occur at six contact
regions in the S-CVT, and the total amount of spin loss reaches almost 0.076 N m.
Spin moments decrease as the variator angle increases, except for the input discs
whose spin moments remain constant (because there are no speed changes in the
input discs). The gross spin loss decreases as the input speed rises, that is due to
the characteristics of our friction model: the friction force at zero relative speed has
the maximal value.
68
0 10 20 30 40 50 60 70
0
2000
4000
6000
8000
decreasing as the
input speed increases
0.020
0.025
0.07
0.08
Figure 4.9: Spin losses on S-CVT at input speed of 3000 rpm.
The average value of spin loss of our numerical results is almost 0.072 N m.
Considering the input torque is limited under the static friction torque of 1.962 Nm,
the ratio of spin loss to static friction torque is almost 3.67%. Considering normal
operating conditions, at which the input torque is smaller than the limiting torque,
one can note that the ratio of spin loss becomes much greater. To reduce this loss,
it is helpful to operate the S-CVT with high input speeds; the increased relative
velocity reduces the relevant friction force.
69
4.4 Slip Motion of the S-CVT
4.4.1 Stick and Slip States
When there are stick states at all contact points of the S-CVT, the exerted tangential
force (e.g., driving, load, or shifting force) must be smaller than the static friction
force. When leaving this stick condition, for example when a certain tangential force
becomes large enough to cause slip in the S-CVT, the relative velocity grows, and
the transmitted force is limited by the kinetic friction force as Equation (4.5).
The dynamics of the S-CVT forms a set of second-order dierential equations
as shown in Equation (2.10). However, when slippage occurs at any contact point
of the S-CVT, the dynamic motion of the S-CVT is determined mainly by friction
forces; therefore the whole equations of motion change.
4.4.2 Slip Loss of the S-CVT
Slip loss is caused by rotational slippage at the contact points, mainly by changes
in the transmitted forces. Thus slip loss T
slip
can be dened as the torque dierence
between the driver and the driven (in the case of the S-CVT, the sphere and discs):
T
slip
= R
driver
F
driver
R
driven
F
f
(4.17)
where R
driver
, R
driven
are the eective contact radii of the driver and the driven
respectively, F
driver
is the driving force, and F
f
is a kinetic friction force.
4.4.3 Slip Involved Contact Analysis
The contact analysis of rotating bodies in rolling conditions were discussed in Section
4.3. The results of that section are based on the assumption that the contact center
does not dislocate from the original contact center. However, when slippage in the
70
rolling direction is induced, the contact center must be dislocated, together with
changes in the normal pressure distribution according to the moment equilibrium
condition at the contact point.
When the tangential force becomes large enough to cause slippage (i.e., larger
than the static friction force), there must be a shear force on the contact region
and thus a reactive moment about the center of the sphere (see Figure 4.10). To
satisfy the moment equilibrium about the sphere center by the reactive moment and
tangential shear force, the contact center must be moved by in the rear direction.
The amount of dislocation can be calculated as
=
RF
P
. (4.18)
As a result of this contact center dislocation, the overall normal pressure distri-
bution must be shifted accordingly, which also brings about changes in the friction
force. The change in normal pressure distribution is depicted in Figure 4.11; in the
Figure 4.10: Dislocation of contact center.
71
Figure 4.11: Change of normal pressure distribution in XZ plane.
XZ plane. The normal pressure distribution will be distorted as shown in this g-
ure; p
max
will be larger considering there is no change in the amount of total normal
force. The applied shear force F exposes a boundary between the stick and slip
region of the contact surface. The friction force F
f
can be roughly determined by
the relation F
f
= N; the increased normal pressure causes a larger static friction
force which can resist the applied shear force. Therefore above a certain value of
normal pressure, slippage does not occur.
4.5 Summary
There are two main sources of power loss resulting from slippage in the S-CVT, spin
and slip loss. Spin loss, which is also a main design issues in traction drives, results
from the elastic contact deformation of rotating bodies having dierent rotational
velocities. Slip loss is generated at the contact points between the sphere and discs in
their rolling direction. Once slippage occurs at those contact points, the transmitted
power becomes dierent from the desired value, and the power transmission can even
fail.
72
To analyze the losses resulting from slippage, we rst reviewed previous analyses
of the friction mechanism. We proposed a modied classical friction model that
describes the friction behavior of the S-CVT including Stribeck (i.e., pre-sliding)
eect. We also performed an in-depth study for the velocity elds generated at the
contact regions along with a Hertzian analysis of deection. Hertzian results were
employed to construct the geometric parameters and normal pressure distributions
of the contact surface with respect to elastic and plastic deformations.
With analytic formulations of the relative velocity eld, deection, and friction
mechanism of the S-CVT, we carried out a quantitative analysis of spin loss. As a
result, an explicit model of spin loss was developed. Spin loss is one of the main
design issues in traction and friction drive designs, and our results can provide an
eective means of measuring and predicting spin loss.
We also described some issues related to the slip loss of the S-CVT resulting
from slippage in the rolling direction. When slippage occurs at any contact point
of the S-CVT, the dynamic motion of the S-CVT is determined mainly by friction
forces; therefore the whole equations of motion change. To predict the behavior of
the S-CVT in stick-slip states, it is important of an instantaneous investigation of
those states using information on velocities, accelerations, and forces at the contact
points. Finally we briey described the contact analysis related to slip, when a shear
force resulting from friction occurs at the contact surface.
73
Chapter 5
Shifting Controller Design via
Exact Feedback Linearization
5.1 Introduction
The most important role of the shifting controller for CVTs is the realization of the
target gear ratio, which is directly related to the input/output ratio of power. When
the shifting command for a certain gear ratio is given, the shifting system must be
stabilized so as to realize the demanded gear ratio with the desired performance
(e.g., little shifting eort, short settling time, etc.). The shifting command of a
CVT can be either a nal value or a trajectory of the target gear ratio. According
to the shifting command, the shifting controller design task is denoted as stabilizer
(or regulator) design for the former and tracker (or servo) design for the latter.
In control theory, a basic problem is how to use feedback in order to modify
the original internal dynamics of a controlled plant so as to achieve some prescribed
behavior. In particular, feedback may be used for the purpose of imposing, on the as-
74
sociated closed-loop system, the (unforced) behavior of some prescribed autonomous
linear system. When the plant is modeled as a linear time-invariant system, this is
known as the problem of pole placement, while in the more general case of a non-
linear model, this is known as the problem of feedback linearization (see [78]-[81]).
Feedback linearization is an approach to nonlinear control design which has at-
tracted a great deal of research interest in recent years. The central idea is to
algebraically transform a nonlinear system dynamics into a (fully or partly) linear
one, so that linear control techniques can be applied. This diers entirely from
conventional linearization in that feedback linearization is achieved by exact state
transformations and feedback, rather than by linear approximations of the dynamics
(i.e., Jacobian linearization).
The shifting system of the S-CVT has second-order nonlinear dynamics, and
the original open-loop system reveals unstable characteristics. In order to cancel
nonlinearities of the S-CVT shifting system, and to make it stable and have good
tracking characteristics, we develop a feedback controller based on the exact feed-
back linearization method in this chapter. We rst investigate the instability of the
original shifting system using Lyapunovs indirect method in Section 2. Section 3
briey reviews the dierential geometric preliminaries for the formal description of
the feedback linearization method. In Section 4, we address the input-state feedback
controller design of the S-CVT shifting system. Finally, we investigate the stabi-
lizing and tracking performance of the dedicated shifting controller by numerical
simulation.
75
5.2 Stability Analysis of S-CVT Shifting System
Lyapunovs (indirect) linearization method is involved with the local stability of a
nonlinear system. It is a formalization of the intuition that a nonlinear system should
behave similarly to its linearized approximation for small range motions. Because all
physical systems are inherently nonlinear, Lyapunovs linearization method serves as
the fundamental justication of using linear control techniques in practice, i.e., that
stable design by linear control guarantees the stability of the original physical system
locally.
Theorem 5.1 (Lyapunovs (indirect) linearization method) Let x = 0 be an
equilibrium point for the nonlinear system
x = f (x) (5.1)
where f : T R
n
is continuously dierentiable and T is a neighborhood of the
origin. Let
J =
f
x
(x) [
x=0
Then,
1. The origin is asymptotically stable if Re(
i
) < 0 for all eigenvalues of J.
2. The origin is unstable if Re(
i
) > 0 for one or more of the eigenvalues of J.
For the detailed proof of this theorem, see pp. 127-130 of [80].
To determine the stability of the S-CVT shifting system, we rst restate the
shifting dynamics in Equation (2.10) into state-space form (5.1). Here we replace
the state x
3
(the rotational speed of sphere) by a matrix transformation, because
it does not aect the shifting dynamics. Letting x
1
= , x
2
=

be the states, the
76
corresponding state-space equation is the following second-order state equation:
x
1
= x
2
x
2
=
1
D
a
22
F
s
a
12
(F
i
cos x
1
F
o
sin x
1
)
(5.2)
where
_
_
a
11
a
12
a
21
a
22
_
_
=
_
_
2
(I
a
+I
v
+m
2
)

2
RI
v

2
2
I
v

I
s
R
+ 2
RI
v

2
_
_
, D = a
11
a
22
a
12
a
21
.
Using the trigonometric transformation, i.e.,
a sin x + b cos x =
_
a
2
+ b
2
sin(x + ), = tan
1
(
b
a
)
Equation (5.2) can be written
x
1
= x
2
x
2
=
1
D
a
12
_
F
2
i
+ F
2
o
sin(x
1
) + a
22
F
s

(5.3)
where
= tan
1
(
F
i
F
o
) .
Considering D =
2I
s
(I
a
+ I
v
+ m
2
)
R
+
4RI
v
(I
a
+ m
2
)

3
is always larger than
zero, the equilibrium point is given by
x

1
= = tan
1
F
i
F
o
, x

2
= 0, F

s
= 0 . (5.4)
We can say that the equilibrium point of interest is x

= (, 0). Physically, this

point corresponds to the steady state of the shifting system in which shifting does
not occur.
The Jacobian matrix J of the shifting system (5.3) linearized about the equilib-
rium point becomes
J =
_
_
0 1
a
12
D
_
F
2
i
+ F
2
o
0
_
_
(5.5)
77
the eigenvalues of J are
i
=
_
a
12
D
_
F
2
i
+ F
2
o
. Hence the linearized system is
unstable, and therefore so is the shifting system of the S-CVT at this equilibrium
point.
Physically this means that when the input and output force relation (F
i
cos =
F
o
sin ) is broken (i.e., steady state is destroyed) by some disturbances from the
input or output force, it must be followed by a change in variator angle (gear ratio)
from the shifting actuator, or by a change in input force from the power source
controller. In order to make the shifting system stable, one can conclude that an
appropriate feedback controller is necessary. In the following sections, we will discuss
the design of a feedback controller based on the exact feedback linearization method.
5.3 Dierential Geometric Preliminaries
We start by recalling some dierential geometric preliminaries (see [78]-[81]); we
then apply these tools to the input-state linearization of our shifting system.
Lie Derivative: Let h : T R be a smooth scalar function, and f : T R
n
be
a smooth vector eld on R
n
. The Lie derivative of h with respect to f or along f ,
written as L
f
h, is dened by
L
f
h = h f .
Repeated Lie derivatives can be dened recursively by
L
0
f
h = h,
L
i
f
h = L
f
(L
i1
f
h) = (L
i1
f
h) f .
Similarly, if g is another vector eld, then the scalar function L
g
L
f
h is
L
g
L
f
h = (L
f
h) g .
78
Lie Bracket: Let f and g be two vector elds on R
n
. The Lie bracket of f and g
is a third vector eld dened by
[f , g] = g f f g .
The Lie bracket [f , g] is commonly written as ad
f
g, where ad stands for adjoint.
Repeated Lie brackets can then be dened recursively by
ad
0
f
g = g,
ad
i
f
g =
_
f , ad
i1
f
g

.
Dieomorphism: A mapping T : R
n
R
n
, dened in a region , is called a
dieomorphism if it is smooth, and if its inverse T
1
exists and is smooth. If the
region is the whole space R
n
, then T(x) is called a global dieomorphism. Global
dieomorphisms are rare, and therefore one often looks for local dieomorphism,
i.e., for transformations dened only in a nite neighborhood of a given point. A
dieomorphism can be used to transform a nonlinear system into another nonlinear
system in terms of a new set of states.
Distribution: Let f
1
, f
2
, . . . , f
k
be vector elds on T R
n
. At any xed point
x T, f
1
(x), f
2
(x), . . . , f
k
(x) are vectors in R
n
and
(x) = spanf
1
(x), f
2
(x), . . . , f
k
(x)
is a subspace of R
n
. To each point x R
n
, we assign a subspace (x). We will
refer to this assignment by
(x) = spanf
1
, f
2
, . . . , f
k

which we call a distribution.

79
Involutive Distribution: A distribution is involutive if
g
1
and g
2
[g
1
, g
2
] .
If is a nonsingular distribution on T, generated by f
1
, f
2
, . . . , f
r
, then it can be
veried that is involutive if and only if
[f
i
, f
j
] , 1 i, j r .
Theorem 5.2 (Frobenius Theorem) Let f
1
, f
2
, . . . , f
r
be a set of linearly inde-
pendent vector elds. The set (equivalently, a nonsingular distribution) is completely
integrable if and only if it is involutive.
Consider an ane nonlinear single input system
x = f (x) +g(x) u . (5.6)
With this system, the input-state linearization problem can be stated as follows:
Find u = (x) + (x) and z = T(x)
such that
z
1
= z
2
,
z
2
= z
3
,
.
.
.
z
n
= .
(5.7)
The following theorem provides a denite criteria for the existence of the input-
state linearization solution and constitutes one of the most fundamental results of
feedback linearization theory.
Theorem 5.3 (Input-State Linearizable Condition of Single Input System)
The nonlinear system (5.6), with f (x) and g(x) being smooth vector elds in R
n
, is
80
input-state linearizable if and only if there exists a region such that the following
conditions hold:
1. the vector elds g, ad
f
g, . . . , ad
n1
f
g are linearly independent in ,
2. the set g, ad
f
g, . . . , ad
n2
f
g is involutive in .
For the detailed proof of this theorem, see pp. 568 of [80] and pp. 239-241 of [81].
The proof of this theorem leads to important relations that can be deduced from the
independent conditions of Lie brackets, which suggests an implicit way of obtaining
an appropriate dieomorphism z = T(x) as follows:
z
1
ad
k
f
g = 0 k = 0, 1, . . . , n 2 ,
z
1
ad
n1
f
g ,= 0 .
(5.8)
Moreover, recursive application of the Lie bracket to the z
n
equation yields
(x) =
L
n
f
z
1
L
g
L
n1
f
z
1
, (x) =
1
L
g
L
n1
f
z
1
.
(5.9)
5.4 Shifting Controller Design via Input-State Lineariza-
tion
Based on the above dierential geometric denitions and theorems, input-state lin-
earization of the shifting system of the S-CVT has been performed via the following
steps:
1. Construct the vector elds g, ad
f
g, . . . , ad
n1
f
g for our system.
2. Check the controllable and involutive conditions.
3. Find the rst new state z
1
from Equation (5.8).
4. Compute the dieomorphism that transforms the state x into the new state z,
T(x) =
_
z
1
L
f
z
1
. . . L
n1
f
z
1

T
, and the input transformation using Equation (5.9).
81
First, we put the shifting system dynamics into the ane nonlinear control sys-
tem form (5.5) in order to obtain the corresponding vector elds f and g. Here we
consider the shifting force F
s
to be the control input u. Then f and g of the shifting
dynamics can be written
f =
_
x
2
a
12
D
_
F
2
i
+ F
2
o
sin(x
1
)
_
T
, g =
_
0
a
22
D
_
T
. (5.10)
Knowing that the system order n = 2 and g = 0, the corresponding Lie bracket
then becomes
ad
f
g = g f f g
= 0
_
_
0 1
a
12
D
_
F
2
i
+ F
2
o
cos(x
1
) 0
_
_
_
_
0
a
22
D
_
_
=
_

a
22
D
0
_
T
.
(5.11)
5.4.1 Controllability and Linearizability
We say the system is controllable if, using appropriate control inputs, the states can
be moved in any direction in the state space. For a linear system such as
x = Ax +Bu
controllability is a property of the pair (A, B) and can be checked as follows.
The pair (A, B) is controllable if and only if the rank of controllability matrix,
C, is n (n is the system order, i.e., dimension of A), where C is given by
C =
_
B AB A
2
B . . . A
n1
B

To determine the controllability of nonlinear systems of the form (5.5), the control-
lability matrix C in the linear system is replaced by
_
g ad
f
g ad
2
f
g . . . ad
n1
f
g

(5.12)
82
In order to determine the controllability of the shifting system of the S-CVT,
we investigate the rank of the controllability matrix using the results of Equations
(5.10) and (5.11):
rank
_
_
0
a
22
D
a
22
D
0
_
_
= 2.
Hence, we can say that the shifting system of the S-CVT is controllable. Further-
more, since the vector elds g, ad
f
g are constant (i.e., its Lie derivatives are zero),
they form an involutive set. Therefore the shifting system is input-state linearizable.
5.4.2 Input-State Linearization
Now we are ready to perform input-state linearization with the new states. First we
nd a dieomorphism T(x)that can transform the original shifting dynamics into
the linearized system. Using the results of Equation (5.8), the necessary conditions
for the rst state z
1
are
z
1
x
1
,= 0,
z
1
x
2
= 0 .
Thus z
1
must be a function of x
1
only. Among the various candidates for z
1
, the
simplest solution is z
1
= x
1
. The other state can be obtained from z
1
z
2
= z
1
f = x
2
.
The corresponding dieomorphism T(x) can be obtained as
z = T(x) =
_
_
x
1

x
2
_
_
. (5.13)
Accordingly the input transformation in Equation (5.9) is
u =
z
2
f
z
2
g
83
which can be written explicitly as
u =
D
a
22

a
12
D
_
F
2
i
+ F
2
o
sin(x
1
) . (5.14)
As a result of the above state and input transformations, we end up with the fol-
lowing set of linear equations
z
1
= z
2
, z
2
= (5.15)
=
a
12
D
_
F
2
i
+ F
2
o
sin(x
1
) +
a
22
D
u . (5.16)
thus completing the input-state linearization.
5.5 Shifting Controller Design
By the above input-state linearization results, we now perform the shifting controller
design which can stabilize the shifting system according to the shift command and
track the demanded variator angle trajectory.
Stabilizing Controller Design
Since the new dynamics (5.15) is linear and controllable, it is well known that the
linear state feedback control law
= k
1
z
1
k
2
z
2
can guarantee asymptotic stability by selecting feedback gains k
1
and k
2
so as to
satisfy the Hurwitz condition. The linearized system can be written
z
1
+ k
2
z
1
+ k
1
z
1
= 0 . (5.17)
84
Tracking Controller Design
For the case of the tracking problem, it is desired to have the variator angle track
a prescribed trajectory
d
. Then the input is designed as
= z
1d
k
1
e k
2
e (5.18)
where e = z
1
z
1d
and z
1d
=
d
. Therefore, the tracking problem of linearized
shifting dynamics transforms into the following error dynamics:
e + k
2
e + k
1
e = 0 . (5.19)
In order to guarantee asymptotic tracking performance of the shifting system, one
may check whether the gain selection k
1
, k
2
can satisfy the Hurwitz condition, sim-
ilarly to the case of stabilizer design. The Hurwitz condition, however, oers only a
set of inequalities for the feedback gains, and the gain selection within these bound-
aries must be achieved using other criteria.
Gain Selection
The resulting closed-loop dynamics of the shifting system (5.17), (5.19) can be
viewed as the canonical form of a general second-order oscillation problem:
s
2
+ 2
n
s +
n
2
= 0 .
Hence one can give physical meaning to the feedback gains as the respective damping
ratio and the natural frequency
n
. The relation between the feedback gains and
,
n
are simply
k
1
=
n
2
, k
2
= 2
n
. (5.20)
85
Case A k
1
= 100, k
2
= 20
Case B k
1
= 50, k
2
= 10

2
Table 5.1: Candidates for k
1
, k
2
.
Therefore, we can deduce the relation of the gains k
1
, k
2
as follows:
k
2
= 2
_
k
1
. (5.21)
Based on previous well-known research results [82], [83] on the vibration of
second-order systems, we consider two cases of k
1
, k
2
(see Table 5.1). In this study,
we desire our shifting controller to provide the most rapid response according to the
shifting command without overshoot; we designate the settling time of the shifting
system (the time in reaching the new equilibrium state) to be less than 1 second.
Hence, we select the system damping ratio to 1, which corresponds to the case
of critical damping. For a given initial excitation, a critically damped system tends
to approach the equilibrium position the fastest without any overshoot. Moreover,
these feedback gains guarantee the asymptotic stability and tracking performance
of the S-CVT shifting system.
Numerical Results
We rst investigate the stability of the shifting system with the proposed feedback
gains. To do this, we simulate the behaviors of the shifting system numerically.
For the simulation conditions, we set the initial states of the system to
= 30

,

= 0, F
i
= 1, F
o
=

3 .
This initial condition is one of the equilibrium points of the shifting system. At
this instant, however, the output force suddenly changes from

3 to 1. Thus the
86
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
30
33
36
39
42
45

(a) Variator angle.
0.0 0.2 0.4 0.6 0.8 1.0 1.2
0.0
0.2
0.4
0.6
0.8
F
s
(b) Control.
Figure 5.1: Stability of the S-CVT shifting system.
input-output force equilibrium no longer holds, and the gear ratio (i.e., the variator
angle ) of the S-CVT must be changed into a new equilibrium state which makes
the system stable. Figure 5.1 shows the numerical results of the system behavior
and corresponding control from Equations (5.14), (5.17). As expected, both cases
of feedback gains show the asymptotic stability of the system. The variator angles
for each case change from the initial state into the new equilibrium point = 45

.
Next we investigate the tracking performance of the shifting system as follows.
For the reference trajectory of the variator angle, we consider a sinusoidal function

3
sin(

2
t ) (see Figure 5.2 (a)), with the initial states of the system chosen as
= 45

,

= 0, F
i
= 1, F
o
= 1 .
Maintaining the input-output force as the initial values, the calculated variator angle
changes are depicted in Figure 5.2 (b) using Equation (5.18). As expected, both
cases of feedback gains show asymptotic convergence of tracking error. The relevant
87
0 2 4 6 8 10
-60
-40
-20
0
20
40
60
(a) Reference shifting command
d
.
0.0 0.3 0.6 0.9 1.2 1.5 1.8
-60
-40
-20
0
20
40
60
0.0 0.2 0.4 0.6 0.8
40
45
50
55
60
(b) Variator angle changes.
Figure 5.2: Tracking performance of the S-CVT shifting system.
tracking error and corresponding shifting eort are shown in Figure 5.3.
For both cases, the system responses match our predened performance measure.
0.0 0.2 0.4 0.6 0.8 1.0 1.2
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0.00
(a) Tracking error.
0 2 4 6 8 10
-0.2
0.0
0.2
0.4
0.6
F
s
(b) Control.
Figure 5.3: Tracking error and corresponding control.
88
From the numerical results, we select the feedback gains for case B, although the
shifting response for case A is faster than that for case B (the time to reach the new
equilibrium variator angle 45

in case A is almost 0.7 second, while for case B it is

almost 1.0 second). The shifting eort (i.e., control eort) for case B maintains a
small value and varies monotonically compared to case A.
Using the selected feedback gains, we reconsider the stability of the system.
The overall system behavior during a gear ratio change is determined from the
S-CVT dynamics (2.10). The rotational speeds of the input, output, and sphere
are depicted in Figure 5.4, using the initial condition and corresponding control in
stability analysis.
0.0 0.2 0.4 0.6 0.8 1.0 1.2
900
1200
1500
1800
2100
2400
2700
Figure 5.4: System behaviors of S-CVT during the gear ratio change.
89
5.6 Summary
Due to the nonlinearity of the S-CVT shifting dynamics, the original open-loop
system is inherently unstable. Hence a feedback controller is necessary to make the
system stable and to achieve eective tracking performance. To do this, we designed
a feedback controller that cancels nonlinearities and transforms the original nonlinear
system dynamics into a stable and controllable linear one, based on the input-state
linearization method.
In this chapter, we showed the instability of the original S-CVT shifting system
using Lyapunovs linearization method. We also briey reviewed the mathemati-
cal background for the formal description of the input-state linearization method.
With this background, we performed the input-state linearization of our system,
and designed a feedback controller which achieves asymptotic stability and eec-
tive tracking performance of the S-CVT shifting system. In selecting the feedback
gains of the proposed controller, we considered our linearized shifting dynamics as a
canonical second-order oscillation problem. In order to achieve a predened shifting
performance, we then set the feedback gains; comparing the numerical results of
the shifting eort (i.e., control eort) and the settling time. Finally we presented
numerical results which demonstrate shifting controller performance with respect to
stability and tracking.
90
Chapter 6
Optimal Control of an S-CVT
equipped Power Transmission
6.1 Introduction
Among the various advantages of a CVT, the most prominent is its ability to run the
power source at the power ecient regime. Furthermore, in most power sources such
as internal combustion engines and electric motors, optimal eciency lies at a certain
operating point. As reviewed in Chapter 1, many control engineers endeavor to nd
an eective way of controlling a CVTs gear ratio to maintain the power source at
the most ecient point and realize shifting commands in the desired manner.
Optimal control of a CVT equipped power transmission is then dened as the
problem of nding a gear ratio which can minimize the energy consumption of the
power source without any losses in output performance. Hence, in order to design
a minimum energy control law for a CVT, one must rst investigate the eciency
characteristics of the power source as well as dene the target performance. Driven
91
mainly by automotive engineers, various control approaches have been tried and
realized. There are two major issues in controlling CVTs to achieve eciency and
performance objectives: power source consolidated control, and establishing the
shifting map (the variogram), which is a look-up table of the speed relations between
the power source and output.
The S-CVT is intended for use in small power capacity power transmissions; thus
a dc motor is considered as the power source in this study. DC motors are designed
to be very ecient at their rated speeds, and it is now generally believed that there
is very little room for improvement in terms of hardware performance. With recent
advances in power electronics, the motor drivers that supply the input voltage or
current are also now extremely ecient, compared to previously used analog drivers,
enough to be used as variable speed drives [84], [85]. In addition, dc motor optimal
control algorithms that take into account the load and other operating characteristics
have been developed [86], [87], further reducing overall power consumption.
In this chapter, we present a minimum energy control law for the S-CVT con-
nected to a dc motor. To do this, in Section 2 we rst describe the general power
eciency characteristics of a dc motor using the well-known dc motor dynamic equa-
tions. In Section 3 we present the results of a numerical investigation of the possi-
bility of energy saving using the S-CVT benchmarked against a standard reduction
gear, taking into account the equations of motion of the S-CVT equipped power
transmission system and an ideal motor model. In addition, a computed torque
control algorithm for the S-CVT is proposed. Section 4 deals with the minimum
energy control design via a B-spline parameterization of the trajectories. Finally we
show some numerical results of energy savings using the proposed minimum energy
control law.
92
6.2 Power Eciency of a DC Motor
6.2.1 DC Motor Dynamics
We now consider a general armature-controlled dc motor as shown in Figure 6.1, in
which the eld current is held constant. We adopt the following nomenclature:
R
a
= armature resistance [ohm],
L
a
= armature inductance [henry],
i
a
= armature current [ampere],
i
f
= eld current [ampere],
e
a
= applied armature voltage [volt],
e
b
= back-emf (electromotive force) [volt],

M
= angular velocity of the motor [rad/sec],

o
= angular velocity of the CVT output shaft [rad/sec],
I
eq
= equivalent moment of inertia of the motor and load referred to
the motor shaft [kg m
2
],
T
M
= motor torque [N m],
T
load
= load torque [N m].
Figure 6.1: Diagram of an armature-controlled dc motor.
93
Then the circuit equation is
L
a
di
a
dt
+ R
a
i
a
+ e
b
= e
a
. (6.1)
The induced voltage e
b
is directly proportional to the speed of the motor
M
, or
e
b
= k
e

M
(6.2)
where k
e
is a back emf constant. The motor torque T
M
is directly proportional to
the armature current:
T
M
= ki
a
(6.3)
where k is a motor-torque constant.
Referring to Figure 6.1, we consider an S-CVT equipped power transmission.
Using a gear train including CVTs at the motor shaft has the eect of reducing not
only the load torque by the gear ratio, but also the equivalent inertia by a square of
the gear ratio; the motor dynamic equation becomes
I
eq

2
d
M
dt
= T
M

T
Load

,
o
=
M
1

(6.4)
where represents the reduction gear ratio. In the case of the S-CVT, however,
is replaced by the torque ratio of the S-CVT, i.e., = cot , so that the motor
dynamic equation becomes
I
eq
tan
2

d
M
dt
= T
M
T
Load
tan ,
o
=
M
tan . (6.5)
Assuming that the armature inductance L
a
in the circuit is small enough to
neglect, we obtain the following equation from (6.1):
R
a
i
a
+ e
b
= e
a
94
thus the motor torque can be written as
T
M
=
k(e
a
k
e

M
)
R
a
. (6.6)
Finally, the dierential equation for the speed of the output shaft
o
(6.5) becomes
I
eq
d
o
dt
+
kk
e
R
a

o
cot
2
=
ke
a
R
a
cot T
Load
. (6.7)
6.2.2 Power Eciency of a DC Motor
In this section, we consider the eciency characteristics of a general armature-
controlled dc motor in Figure 6.1. The torque produced by a dc motor is directly
proportional to the armature current (6.3); when the equivalent inertia and/or the
load torque applied at the motor shaft is increased, the armature current must also
be increased. Rearranging the above equations and using the fact that the value
of k
e
is equal to k, the relationship between the mechanical power and the electric
power is
T
M

M
= e
a
i
a
i
a
2
R
a
. (6.8)
In the above equation, the i
a
2
R
a
term represents the electric power-loss, called the
armature-winding loss, generally dissipated through heat generation.
From Equation (6.1), at certain values of the armature voltage, decreasing the
armature current will increase the value of the back-emf and the motor speed. Based
on these observations, one can notice that motors have their highest eciency in the
low-torque, high-speed region. Figure 6.2 depicts the power eciency of a general
dc motor with respect to the motor speed and the load torque. As can be seen from
the graph, motor eciency is highest in the low-torque, high-speed region (indicated
in dark blue), with the eciency dropping o steeply in other regimes. To enhance
95
Figure 6.2: Eciency of an armature-controlled dc motor.
the power eciency of a dc motor, it is clearly advantageous to operate it in this
region of maximum eciency.
6.3 Investigation of S-CVT Energy Savings
In this section, we perform a numerical investigation of the energy savings of the
S-CVT. As a comparative benchmark, we consider two power transmission systems
driven by the same dc motor: one driven by a reduction gear, the other driven by
the S-CVT. A typical output speed prole is given to each system, and then the
variations of electricity (e.g., ampere and voltage) are calculated. In the case of a
reduction gear, the motor speed is controlled in order to follow the given output
speed prole. However, in the case of the S-CVT, one can control either the gear
ratio (i.e., the variator angle) or the motor speed. In this study we manipulate the
96
Rated voltage 12 V olts
Rated power 60 Watts
Motor-torque constant 0.0272 N m/A
Back emf constant 0.0272 volt sec/rad
Rotor winding resistance 0.48 Ohm
Stall torque 0.68 N m
Table 6.1: Characteristic coecients of dc motor.
output speed by the variator angle only, choosing to operate the motor at its most
ecient regime by eectively treating the armature voltage e
a
as its rated value.
For the numerical investigation, we assign a load torque T
Load
of 0.07 Nm, and
an equivalent inertia with respect to the input shaft I
eq
of 0.01 kgm
2
; for these
values, the stall torque is calculated to be 1.6 Nm. We therefore choose a dc motor
with a power rating of 60 Watts (see Table 6.1 for the detailed specications of the
dc motor) and a gear ratio of four for the reduction gear case. The desired output
speed prole is chosen to be a sinusoid with a magnitude of 500 rpm and a period
of 40 seconds (See Figure 6.3).
0 10 20 30 40
-600
-400
-200
0
200
400
600
Figure 6.3: Target prole of output speed.
97
6.3.1 Control Design based on the Computed Torque Method
In this section, we design a control based on the computed torque control method
which is used widely in robotics and other engineering elds. The computed torque
control compensates for tracking errors by using feedback information about the dif-
ferences between the predened objective trajectories of position, speed, acceleration
and the estimated actual trajectories.
From the given output speed prole
o
(t), the computed motor torque T
M
is
calculated using Equation (6.6) and the relation of
o
between
M
:
T
M
=
k
R
a
(e
a
k
e

o
cot ) .
This motor torque must be balanced by the torque due to the equivalent inertia and
the load (6.5), i.e.,
k
R
a
(e
a
k
e

o
cot ) = (I
eq
d
o
dt
+ T
Load
) tan .
Rearranging the above equation yields
kk
e
R
a

o
cot
2

ke
a
R
a
cot + I
eq
d
o
dt
+ T
Load
= 0 . (6.9)
Hence the variator angle prole can be obtained by solving the second-order poly-
nomial Equation (6.9). Given the desired output speed prole
o
(t) as shown in
Figure 6.3, and assuming a xed armature voltage of 12 V olts, we determine the
trajectory of the variator angle (t) (see Figure 6.4).
Comparing the results with the sinusoidal shape of the target output speed prole
in Figure 6.3, the variator angle time prole is seen to have a sharper gradient during
the rise stage. This dierence can be accounted for by the acceleration rate of the
output speed. In general reduction gear equipped transmissions, the input speed (the
motor speed here) varies directly with the output speed; the required acceleration
98
0 10 20 30 40
-8
-6
-4
-2
0
2
4
6
8
Figure 6.4: Computed variator angle time prole.
rate of the output speed is generated entirely by the input torque (in this case the
motor torque). However, CVTs can manipulate the gear ratio according to the
output speed. Therefore the variator angle prole of the S-CVT is aected by the
shape of the desired output speed prole and the necessary acceleration rate. The
magnitude of the angle variation, which for our case can be regarded as the control
eort, is small enough such that it can be implemented even with a relatively ecient
small-capacity shifting actuator.
6.3.2 Numerical Results
Using the equations derived previously, we now perform a numerical study of the
energy consumption rates for each system. In Figure 6.5 (a), the initial motor speed
of the S-CVT system is set to about 4200 rpm, which is obtained from a zero-load
condition for the dc motor considered here. While the motor speed for the reduction
gear case varies depending on the output speed prole, for the S-CVT it remains
close to 4000 rpm which is nearly the nominal speed for the zero-load condition. The
99
0 10 20 30 40
-2000
-1000
0
1000
2000
3000
4000
5000
(a) Motor speed.
0 10 20 30 40
-0.30
-0.15
0.00
0.15
0.30
0.45
(b) Motor torque.
Figure 6.5: Motor behaviors; reduction gear vs. S-CVT.
torque exerted on the motor for each case is calculated in Figure 6.5 (b). Generally,
maximal torque is necessary when the motor rst begins to rotate. However, for the
S-CVT, almost zero torque is exerted when the motor starts rotating. This reduction
in load torque is a consequence of the innite torque multiplication characteristics
of the S-CVT, which can be realized in the vicinity of zero variator angle.
0 10 20 30 40
0
20
40
60
80
100
120
140
Figure 6.6: Power consumption; reduction gear vs. S-CVT.
100
Reduction gear S-CVT
1783.2 Joules 957.4 Joules
Table 6.2: Energy consumption; reduction gear vs. S-CVT.
The consumed energy for each case is calculated using the above results and the
relation
Energy =
_
[ e
a
(t) i
a
(t) [ dt.
We regard negative values of current and voltage as part of the overall consumed
energy. From Figure 6.6, we have calculated the energy consumption in Table 6.2
consequently. Our results suggest that in principle, the consumed energy for the
case with the S-CVT is less than the other case by almost 46.3%, although in actual
implementations the eects of friction, backlash, and other sources of loss will have
to be considered in more detail.
6.4 Minimum Energy Control via a B-Spline Parame-
terization
In this section, we describe the minimum energy control design via a B-Spline pa-
rameterization. The minimum energy control problem of an S-CVT equipped power
transmission is dened as follows:
Find the optimal control u that minimizes J =
_
t
f
t
0
i
a
e
a
dt
subject to I
eq

o
=
kk
e
R
a

o
u
2
+
ke
a
R
a
u T
Load
(6.10)
where u is the gear ratio, i.e., cot . The boundary conditions can be expressed
in various forms according to the target performance. In this section, we consider
101
the more complicated case of the S-CVT application for some position changers,
e.g., a mobile robot, a vehicle, a positioning table, etc. For this case, the boundary
conditions are given as follows:

o
(t
0
) =
o
(t
f
) = 0, s(t
0
) = 0, s(t
f
) = d
where d is the desired displacement, s(t) represents the displacement prole, and t
o
,
t
f
represent the initial and nal times respectively.
6.4.1 B-Spline Parameterization
A solution to the above optimal control can be found by assuming that the displace-
ment prole s(t) is parameterized by a B-spline. The B-spline curve depends on
the basis functions B
i
(t) and the control points p = [p
1
. . . p
n
] with p
i
R . The
displacement prole then has the form s = s(t, p) with
s(t, p) =
n

i=1
B
i
(t)p
i
(6.11)
Using this formulation (6.11),
o
,
o
and u, which are functions of t and p, can
be written as

o
(t, p) =
1
r

t
s(t, p),
o
(t, p) =
1
r

2
t
2
s(t, p),
and
u(t, p) =
ke
a
+
_
D(t, p)
2kk
e

o
(6.12)
where D(t, p) = k
2
e
2
a
4R
a
kk
e

o
(T
Load
I
eq

o
), r is the conversion factor from a
rotational speed into a linear speed (for the cases of mobile robots and vehicles this
means the wheel radius). The control u is determined from Equation (6.12), but an
102
additional inequality constraint D(t, P) 0, t [t
0
, t
f
] must also be satised. In
order to satisfy the boundary conditions, we set p
1
, p
2
to zero and p
n1
, p
n
to d.
Setting the input voltage e
a
to a constant value by the same reason as in the
previous section, the armature current i
a
can be calculated from Equations (6.5)
and (6.6):
i
a
(t, p) =
e
a
k
e

o
u
R
a
Hence the original optimal control problem is converted into a parameter optimiza-
tion problem as follows:
minimize J(p) = e
a
_
t
f
t
0
i
a
(t, p)dt
subject to D(t, p) 0 , t [t
0
, t
f
] .
(6.13)
6.4.2 Gradients of the Objective Function and Constraint
To apply various parameter optimization algorithms (i.e., steepest descent, modied
Newton method, quasi-Newton method, penalty method, etc.) to this problem
(6.13), we must formulate the gradients of the objective function and constraint
because almost all optimization algorithms require gradients of the objective function
and constraint.
The gradient of the objective function is
J
p
i
=
_
t
f
t
0
i
a
p
i
e
a
dt
where the partial derivatives of i
a
are as follows:
i
a
p
i
=
k
e
R
a
_

o
p
i
u +
o
u
p
i
_
The derivatives of
o
and u are obtained from the fact that
s
p
i
= B
i
(t)
103
Since the constraint in Equation (6.13) is represented in the form of inequality,
we can just know whether the constraint is eective or ineective. In order to nd
the gradient of the constraint, we now propose the new constraint by dening a new
function, g(t, p), g(p), as follows:
g(t, p) =
_

_
D(t, p) if D(t, p) < 0
0 if D(t, p) 0
g(p) =
_
t
f
t
0
g(t, p)dt .
Figure 6.7 illustrates the interpretation of g(p). With these denitions, it is apparent
that the constraint in Equation (6.13) is equivalent to the following constraint:
g(p) = 0
It is dicult to solve this constraint analytically; however the gradient is well dened.

Figure 6.7: Interpretation of g(p).

104
g(p) can be redened as
g(p) =

j
_

j

j
D(t, p)dt
where D(t, p) < 0 for t [
j
,
j
] and D(
j
, p) = D(
j
, p) = 0. The gradient of
g(p) can be dened as follows:
g
p
i
=

p
i
_

j

j
D(t, p)dt
=

j
_
_

j

D(t, p)
p
i
dt D(
j
, p)

j
p
i
+ D(
j
, p)

j
p
i
_
=

j
_

j

D(t, p)
p
i
dt .
The gradient is now rewritten as follows:
g
p
i
=
_

D(t, p)
p
i
if D(t, p) < 0
0 if D(t, p) 0
, (6.14)
g(p)
p
i
=
_
t
f
t
0
g(t, p)
p
i
. (6.15)
Because it is dicult to nd
j
,
j
for a given p, we can alternatively use Equation
(6.15) to numerically calculate the gradient of the constraint.
6.4.3 Numerical Results
We determine by simulation the power consumption for a minimum energy control
and a comparative computed torque control (see more detail in Section 6.3). The
comparative control is designed to manipulate the output speed in a sinusoidal
105
fashion, satisfying the boundary conditions. To satisfy the boundary conditions the
displacement prole is described as follows:
s(t) =
d
2
_
1 cos
t
t
f
_
From this relation,
o
,
o
are derived by dierentiation, and the control u is obtained
from Equation (6.9).
Based on the mathematical models presented in the previous sections, we have
developed a simulation program with MATLAB. This program uses Simpsons rule
for integration and the BFGS quasi-Newton method for optimization. We assign
the nal time t
f
to be 5 seconds, and the desired displacement d to be 8 meters.
Figure 6.8 depicts the corresponding variator angle time trajectories which are
directly related with the controls for each case. In this gure, the optimized variator
angle is much atter than in the case of the computed torque control. The resulting
motor speed and torque are calculated in Figure 6.9. As can be seen in Figure 6.9,
in the minimum energy control case the variation of the motor speed is smaller and
0 1 2 3 4 5
0
5
10
15
20


Figure 6.8: Optimal variator angle time prole.
106
0 1 2 3 4 5
3000
3200
3400
3600
3800
4000
4200
4400
4600
4800
5000

(a) Motor speed.
0 1 2 3 4 5
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20


(b) Motor torque.
Figure 6.9: Motor behaviors with the minimum energy control.
the motor torque is closer to zero compared to the other controller.
Figure 6.10 shows the output behavior of the S-CVT equipped power transmis-
sion. From this gure, one can see that the minimum energy controller accelerates
the output faster than the computed torque control. Consequently, we have calcu-
0 1 2 3 4 5
0
2
4
6
8

(a) Displacement.
0 1 2 3 4 5
0.0
0.5
1.0
1.5
2.0
2.5


(b) Output speed.
Figure 6.10: Output behaviors with the minimum energy control.
107
Minimum energy control Computed torque control
43.37 Joules 57.16 Joules
Table 6.3: Energy consumption with the minimum energy control.
lated the energy consumption in Table 6.3. The optimized energy consumption is
less than that of the other case by almost 24.1%
6.5 Summary
Using an ideal motor model, we carried out a numerical study on the energy eciency
of an S-CVT equipped power transmission system, and compared the results with
that of a standard reduction gear. The S-CVT was intended to primarily for use
in small power capacity transmissions, thus a dc motor was considered here as the
power source. In this chapter, we presented a minimum energy control law for the
S-CVT in a typical power transmission with a dc motor.
To do this, we rst described the general power eciency characteristics of a dc
motor using well-known dc motor dynamics relations. We then presented the nu-
merical results for the investigation of the S-CVT energy saving possibility, bench-
marked against a standard reduction gear. In addition, we proposed a computed
torque control algorithm for the S-CVT. Section 4 deals with the minimum energy
control design via a B-spline parameterization. By parameterizing the displacement
prole in terms of a B-spline, the optimal control problem is converted into a pa-
rameter optimization problem involving the B-spline control points. Finally, to show
the eectiveness of the developed minimum energy control law, computer simulation
results using a computed torque control and an optimal control law for the same
system are addressed.
108
Chapter 7
Case Study: An S-CVT based
Mobile Robot
We propose an S-CVT based mobile robot (named as MOSTS: Mobile rObot with a
Spherical Transmission System) to put the various advantages of S-CVT including
the originally intended CVT characteristic of energy eciency into practical use.
In this chapter, we rst address the motivation for applying the S-CVT to a
wheeled mobile robot, by rst reviewing the current hardware designs of mobile
robots and their power eciency in Section 1. Section 2 shows the hardware design
of our S-CVT based mobile robot. In addition, we propose a novel pivot mechanism
that uses an internal gear and an uncontrolled dc motor. In Section 3, we perform
numerical simulations and experiments to validate the robots operation, the CVT
characteristics, and its energy saving possibility.
109
7.1 Motivation for Mobile Robot Applications
In recent years, there has been an explosion of research activity in mobile robots,
driven in part by the focus on service robots and their applications, e.g., patient
transportation, autonomous security services, mobile platforms for manipulators,
etc.A large part of the mobile robot literature addresses issues in their planning and
control, taking into account the non-holonomy generated by the wheels. In contrast
to the literature on mobile robot motion planning, relatively little attention has been
given to hardware platforms and other physical aspects of mobile robots [88], [89].
While the mechanical hardware specications for mobile robots vary widely, gen-
erally a wheeled mobile robot requires at least two actuators for moving about in
the plane, each with a dedicated controller (see Table 7.1). Wheel drives, generally
known as dierential drives, are said to be omnidirectional mobile robots as they
can move about arbitrarily in the planar workspace. Track drives, using tracks (or
caterpillars), use at least four track-drive motors with idle track-wheels. Although
track drives can be used in the desert, muddy or undeveloped grounds as they can
move on rough surfaces, owing to the track characteristics they show low power
eciency.
Number of
Drive Motors
Turning
Device
Workspace Features
Wheel
Using
Dierential Gear
1
Steering motor
is necessary
Plane
Controllers
for each motor
Drives Dierential
Drives
Equal to
number of wheels
Unnecessary,
Pivot/Steering
Plane
Controllers
for each motor
Track Drives at least 4
Unnecessary,
Pivot/Steering
Rough
surfaces
Controllers
for each motor,
low eciency
Legged Robots
Equal to
number of joints
Unnecessary
Rough
terrain
Controllers
for each joint
Table 7.1: Hardware specications of general mobile robots.
110
Besides the hardware specication aspects, mobile robots typically use electric
motors as actuators, in particular dc motors, because of their relatively simple con-
trol features, and the fact that power can be supplied from battery sources. Despite
advances in motor eciency, the runtime of mobile robots is still limited by their
batteries and reliance on load conditions. Furthermore, general dc motors have their
best power eciency in the regime of low-torques and high speeds. Thus, it is clear
that the overall power eciency of a mobile robot depends on that of the dc motor
which is adopted in the robot. Consequently to prolong the robots run time, it is
advantageous to operate the motor in this region of maximum eciency.
Hence, to improve the run time and to avoid having to use an oversized motor,
general wheeled mobile robots and vehicles typically use gear reduction [89]. A
reduction gear reduces the load torque and increases the motor speed by the selected
gear ratio. Although automobiles have a nite range of available gear ratios, it is
generally impractical to equip mobile robots with standard transmission devices due
to manufacturing costs, space, and other limitations.
In the case of reduction gears, the motor is operated in the low-eciency region
when accelerating or decelerating the mobile robot due to its xed gear ratio. The
CVT allows for innite ranges of gear ratio, and oers the possibility of much im-
proved energy eciency and performance. Moreover, it allows the motor to deliver
a range of torques at its most ecient speed (the so-called rated speed) while the
mobile robot moves, by changing its gear ratio continuously.
111
7.2 MOSTS: An S-CVT Mobile Robot
7.2.1 Pivot Device for Planar Accessibility
As previously seen in typical mobile robot designs, an additional controlled actua-
tor, such as a steering wheel or a motor for dierentiating each wheel velocity, is
necessary in order to move a mobile robot in the plane. Employing a novel pivot
device, however, we can eliminate the need for an additional steering actuator and
controller. To change its heading direction, MOSTS turns about its center (or piv-
ots) by rotating one of the wheels in the reverse direction. For this to occur, we have
been inspired by the fact that the S-CVT can locate arbitrarily the orientation of
the output shaft. To achieve this, it is necessary to locate one of the output shafts
on the opposite side of the sphere (see Figure 7.1 (a)).
For this operation we have adopted an internal gear driven by a simple actuator
(see Figure 7.1 (b)), e.g., a limit switch used in automated windows, and an uncon-
(a) Pivot inspiration. (b) Realization by use of an internal gear.
Figure 7.1: Pivot device for planar accessibility of MOSTS.
112
trolled motor; this is publicized in the form of patent, pending by the Oce of the
Patent Administration of Korea [90]. Using simple analog devices, we build a pivot
switch that can be turned o according to a pre-set current limit.
The electric circuit diagram of the pivot switch including the driving motor cir-
cuit is shown in Figure 7.2. In the electric circuit of the pivot switch and driving
motor, one can nd that there is no speed controller for the driving motor and pivot
Signal +5V, 0V
Relay 2 level
Setting
M
Driving Motor
-
+
741
-
+
C3117
10K
F m 330
D560
W 47 R
W 5
2 . 0 W
1K
Amp.out
1A=0.1V
Relay2 (12V)
W 5
70W
Relay1
(10V)
+12V
NO
Relay 1
10K
NO
Relay 2
0.5K
Offset Control
C3989
+12V
+12V
5 . 0
1
5 . 0
= =
K
K
Gain
Sig. +5V,0V
4.7K
M
741
-
+
741
-
+
+12V
D560
D560
D560
W 5
2 . 0 W
NO NO
NC NC
C C
C
C
NO
NO
+12
Relay2
Relay3
A B
5K
5K
+12
+12
+12
Offset
1K
18K
Stop level
Control
103
1.4K
Relay1
Relay1 Relay3
C
NO NC
Relay2
Pivot Switch
D 560
+12
10K
F m 220
A
B
Figure 7.2: Electric circuit diagram of pivot switch and driving motor.
113
motor. During pivot motion, each wheel rotates in opposite directions with the
same magnitude, while the driving motor rotates continuously without any changes
of state. The amount of pivot angle is determined by the amount of angular dis-
placements of each wheel, which is controlled by the shifting actuator, or variator.
Moreover, if a controlled actuator is used to rotate the movable output shaft, steer-
ing motion can be obtained. Designed in this fashion, MOSTS has the capability
to move in the plane with one drive motor, one controller for the S-CVT, and one
switching actuator.
7.2.2 Prototype Design
For the construction of the mobile robot platform, we have set the following perfor-
mance targets:
1. A top speed of 5 m/sec;
2. A maximum ascending angle of 10

;
3. A combined vehicle-payload mass of 50 kgs.
To satisfy these goals, we begin by specifying the static-load conditions. The total
resistive force on the wheel is given by
F
resistant
= W sin + C
1
cos
where we assume the drag force coecient C
1
to be 0.75 kgf based on typical
values for the friction coecient between the wheel and the ground, represents
the ascending angle, and W is the weight of the mobile robot platform. The static
friction coecient between the sphere and the output disc is assumed to be 0.12,
while the distance between the disc center and the sphere-disc contact point is set
to be 10 mm. The normal force exerted on the contact point is set to be 80 kgf, or
114
Rated voltage 12 V olts
Rated power 150 Watts
Motor-torque constant 0.0164 N m/A
Back emf constant 0.0164 volt sec/rad
Rotor winding resistance 0.117 Ohm
Stall torque 2.03 N m
Table 7.2: DC motor charateristic coecients of MOSTS.
equivalently 784.8 N. The maximum torque that can be transmitted by the S-CVT
in this case becomes 0.942 Nm.
From the above hardware specications and material properties of the S-CVT,
we choose a specic dc motor that produces a power of 150 Watts with 12 V olts
under nominal operating conditions as the driving motor (see the details provided
in Table 7.2).
The body of the mobile robot is designed to have a cylindrical shape, and a caster
wheel is added to provide stable support. The internal body consists of three layers:
a mechanical base for the transmission system, an intermediate layer for the battery
pack and controller, and a top layer for peripherals and accessories, e.g., navigation
sensors, manipulators. Rotary encoders sensing the speeds of the input and output
shafts are also included. The overall size of the platform is 260 mm in radius, and
500 mm in height (see Figure 7.3).
7.3 Numerical and Experimental Results
In this section, we present numerical and experimental results that demonstrate the
operation of MOSTS, and the energy savings possible from the use of the S-CVT
115
Figure 7.3: Hardware prototype of MOSTS.
N
S
E W
Figure 7.4: The desired trajectory.
116
mechanism over standard reduction gears.
The reference path is shown in Figure 7.4; there are three linear movements
and two pivot motions during 22 seconds. The distance traversed by the robot is 20
meters. During the pivot motion, there is an auxiliary 2 second period for actuating
the pivot switch, which is necessary to move one of the output shafts of the S-CVT
to the opposite direction.
With this reference trajectory, we calculate the necessary wheel velocity prole
satisfying the time constraints by using a sine function (see Figure 7.5). The pivot
motions in the path are specied as a 90

counter-clockwise rotation, followed by a

90

clockwise rotation.
First, we calculate the value of the output speed acceleration from the driving
pattern under the assumption that the input voltage is held constant at 12 V olts.
The exerted load torque is set to 2.5215 Nm, and the equivalent inertia with respect
to the motor shaft is set to 0.01 kgm
2
. With these values and the output speed, we
0 5 10 15 20 25
-100
-50
0
50
100
150
200
250
300
Figure 7.5: Calculated wheel velocity prole.
117
0 5 10 15 20 25
-2
0
2
4
6
8
Figure 7.6: Trajectory of variator angle.
extract the necessary variator angle by a computed torque control algorithm in
Section 6.3. Finally, the trajectory of the variator angle is presented in Figure 7.6.
7.3.1 Numerical Results
Using the equations derived in the previous chapter, we have developed a simulation
program that computes the motor speed, produced torque, and the power consump-
tion. We use the Runge-Kutta fourth-order algorithm for numerical integration in
the simulation program.
In Figure 7.7 (a), the initial motor speed is about 7000 rpm, which is obtained
from the no-load condition of the dc motor considered here. During the whole
operation period, the motor speed varies freely between 6500 rpm and 7000 rpm
regardless of the behavior of the robot (stop, start, and pivot motions), which are
almost the nominal speeds under a no-load condition. The necessary motor torque
is calculated in Figure 7.7 (b). Generally, maximal torques are necessary when the
118
0 5 10 15 20 25
6600
6700
6800
6900
7000
(a) Motor speed.
0 5 10 15 20 25
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
(b) Motor torque.
Figure 7.7: Motor behaviors of MOSTS.
motor of general mobile robots starts rotating. However, in the case of our robot,
almost zero torque is exerted at the start, and the torque variations are quite small
during the whole period.
To investigate the increase in energy savings, we calculate the energy consump-
tion rate of our mobile robot for the reference trajectory. As a benchmark, we
consider a dierential drive type mobile robot having a reduction gear unit with a
gear ratio of six under the same load condition. The dierential drive type robot
considered has two driving motors of 150 Watts at each wheel shaft and follows the
same reference trajectory. Consequently, we calculate the energy consumption rates
for each case using the following equation:
Energy =
_
[ e
a
(t) i
a
(t) [ dt.
From Figure 7.8, we calculate the total energy consumption to be 1389.61 Joules
for the dierential drive with reduction gear unit and 727.86 Joules for our CVT-
119
0 5 10 15 20 25
-10
0
10
20
30
40
50
60
70
80
90
100
110
Figure 7.8: Power consumption; MOSTS vs. dierential drive.
based mobile robot. Our mobile robot equipped with the S-CVT consumes less than
47.6% of the energy consumed by the dierential drive, a signicant improvement
in energy eciency.
7.3.2 Experimental Results
Using the sequential manipulation of the variator angle according to calculated val-
ues of Figure 7.6, we experimentally determined the actual energy consumption of
MOSTS under the same reference trajectory mentioned above. The actual energy
consumption is 1294.92 Joules, which is larger than the ideal case by 567.06 Joules.
However, this is still smaller than the calculated energy consumption of 1389.61
Joules for the dierential drive case (the actual energy consumption for this case
will most likely be signicantly higher than the calculated ideal rate).
To investigate the reason behind this dierence in total energy consumption,
120
0.0 0.5 1.0 1.5 2.0
6200
6300
6400
6500
6600
6700
6800
6900
7000
(a) Motor speed.
0.0 0.5 1.0 1.5 2.0
-1
0
1
2
3
4
5
6
7
8
(b) Motor induced current.
Figure 7.9: Experimental results.
the induced motor current and the actual motor speed for the rst two seconds are
depicted in Figure 7.9. As the reference motion trajectory considered here has ve
repetitive sequences (see Figure 7.5, 7.6, and 7.7), it is sucient to investigate the
rst two second period experimental results. Observe that the initial motor current
is almost 4 Amperes, whereas the ideal value is almost zero. This initial induced
motor current is mainly due to the power loss resulting from manufacturing errors
including bearing friction, gear backlash, etc.Consequently, this power loss makes
the driving motor run at lower speeds, causes the overall power eciency to decrease.
MOSTS
simulation result
experimental result
727.86 Joules
1294.92 Joules
Dierential drive
with reduction gear
simulation result
experimental result
1389.61 Joules
??
Table 7.3: Energy consumption; MOSTS vs. dierential drive.
121
7.4 Summary
In this chapter, we have presented the design of a CVT-based mobile robot using
a minimal number of actuator and control components, by taking advantage of
the typical characteristics of the S-CVT. Such a CVT-based mobile robot has the
advantage of being able to operate the motors in their regions of maximum eciency,
thereby prolonging the total run time of the robot. The addition of a novel pivot
device also enables the mobile robot to achieve steering (more precisely, changing
its heading direction) by using only a single drive motor and controller, unlike most
existing mobile robot platforms.
We also perform an in-depth analysis of the energy eciency of our mobile
robot taking into account features of the dc motors, the S-CVT, and the mobile
robot dynamics. The results are benchmarked numerically with a dierential drive
type mobile robot equipped with a reduction gear. Furthermore, we perform an
experiment using the prototype robot to verify the robots operation and the CVT
characteristics. The numerical and experimental results show that our mobile robot
with S-CVT consumes power less than dierential drive type robots.
122
Chapter 8
Conclusion
In this thesis we have performed a comprehensive study on the design, analysis, and
control of the Spherical CVT. Based on these results, we can conclude that proposed
Spherical CVT shows attractive advantages, such as compact and simple design and
relatively simple control features, eective in particular for mechanical systems in
which excessively large torques are not required and we have performed theoretic
and practical works which could conrm these advantages with a case study of a
Spherical CVT-based mobile robot.
The important conclusions from this work are summarized as follows.
The S-CVT is marked by its simple conguration, innitely variable transmis-
sion (IVT) characteristics and realization of the smooth transitions between
forward, neutral, and reverse states without any brakes or clutches. The power
transmission mechanism is based on dry rolling friction between the contact
bodies of the sphere and discs. Its practical applications are currently lim-
ited to small power capacity mechanical systems, though adopting traction
uid can increase the maximum torque of the S-CVT so as to make its use in
123
traction drive possible.
The S-CVT is intended to overcome some of the limitations of existing CVTs,
e.g., dicult shifting controller design, and the necessity of a large-capacity
and typically inecient shifting actuator. The analysis results on operating
principles, transmission ratios and power eciency of S-CVT have been veried
by experimental results obtained with the testbench.
Spin loss, which is one of the main design issues on traction drives, is analyzed
from its physical mechanism to a quantitative explicit formulation. To analyze
this, we have proposed a modied classical friction model, which can describe
the friction behavior of the S-CVT including pre-sliding eects (i.e., Stribeck
eects). Additionally, we have performed an in-depth study of velocity elds
generated at the contact regions along with a Hertzian analysis of deection.
To stabilize and achieve eective tracking performance we have designed a
feedback controller, which can cancel typical nonlinearities and transform the
original nonlinear system dynamics into a stable and controllable linear one,
based on the input-state linearization method. The designed feedback shifting
controller shows asymptotic stability and tracking performances; the settling
time is smaller than 1 second, the shifting eort varies monotonically and
keeps small value.
Using an ideal motor model, we have presented the numerical results for the
investigation of the S-CVT energy saving possibility, benchmarked against a
standard reduction gear. The minimum energy control design via a B-spline
parameterization is carried out by parameterizing the displacement prole in
terms of B-splines; the original optimal control problem is converted into a
124
parameter optimization problem involving the B-spline control points. To show
the eectiveness of the developed minimum energy control law, simulation
results using a computed torque control and an optimal control law for the
same system are addressed.
We have presented the design of a CVT-based mobile robot using a minimal
number of actuator and control components, by taking advantage of the typi-
cal characteristics of S-CVT. The addition of a novel pivot device also enables
the mobile robot to achieve steering (more precisely, changing its heading di-
rection) by using only a single drive motor and controller, unlike most existing
mobile robot platforms.
The energy eciency of our mobile robot is benchmarked numerically with
a dierential drive type mobile robot equipped with a reduction gear. Fur-
thermore, we perform an experiment using the prototype robot to verify the
realization of robots operation and the CVT characteristics. The numerical
and experimental results show that our mobile robot with S-CVT consumes
the electric power less than that of a dierential drive type robot, signicantly.
125
References
[1] Youngdug Choi and Innwoog Yeo, 1999, Technical Report: Trends of Continu-
ously Variable Transmission, Technical Center, Daewoo Motor Company.
[2] Proc. of Inter. Conf. on Continuously Variable Power Transmissions CVT 96,
Society of Automotive Engineers of Japan, Yokohama, Sept. 11-12, 1996.
[3] Proc. of the 7th Inter. Power Transmission and Gearing Conference, Design
Engineering Division, ASME, San Diego, California, Oct. 6-9, 1996.
[4] Proc. of Inter. Congress on Continuously Variable Power Transmissions CVT
99, Sponsored by I Mech E, JSAE, KIVI, SAE, VDI-EKV, Eindhoven, Sept.
16-17, 1999.
[5] Report of U. S. Department of Energy, 1982, Advanced Automotive Transmis-
sion Development Status and Research Needs, DOE/CS/50286-1.
[6] Philip G. Gott, 1991, Changing Gears: The development of the Automatic
Transmission, Society of Automotive Engineers.
[7] S. H. Loewenthal, 1982, A Historical Perspective on Traction Drives and Re-
lated Technology in Advanced Power Transmission Technology, NASA CP-
2210, pp. 79-108, Fisher, G. K., ed.
126
[8] S. H. Loewenthal, N. E. Anderson, D. A. Rohn, 1983, Advances in Traction
Drive Technology, NASA TM-83397, pp. 1-2.
[9] Tokyo Motor Show, 1999, Tokyo, Japan.
[10] Appletons Cyclopedia of Applied Mechanics, 1880.
[11] N. Moronuki and Y. Furukawa, 1988, On the Design of Precise Feed Mechanism
by Friction Drive, in Journal of the Japan Society for Precision Engineering,
pp. 2113-2118, Vol. 11.
[12] K. Kato, 1990, Fundamentals and Applications of Friction Drive, in Journal
of the Japan Society for Precision Engineering, pp. 1602-1606, Vol. 9.
[13] H. S. Lee and M. Tomizuka, 1996, Robust Motion Controller Design for High-
Accuracy Positioning Systems, in Transactions on Industrial Electronics , pp.
48-55, Vol. 43, No. 1, Feb.
[14] W. S. Chang and K. Y. Toumi, 1998, Modeling of an Omni-Directional High
Precision Friction Drive Positioning Stage, in Proc. of Inter. Conf. on Robotics
and Automation , pp. 175-180, Leuven, Belgium.
[15] M. K. Kurosawa, M. Takahashi, and T. Higuchi, 1998, Elastic Contact Con-
ditions to Optimize Friction Drive of Surface Acoustic Wave Motor, in Trans-
actions on Ultrasonics, Ferroelectrics, and Frequency Control , pp. 1229-1237,
Vol. 45, No. 5, Sep.
[16] R. W. Carson, 1977, 100 years in review: Industrial Traction Drives, in Power
Transmission Design, Oct.
[17] D. Dowson and G. R. Higginson, Elasto-Hydrodynamic Lubrication, Pergamon.
127
[18] J. L. Tevaarwerk and K. L. Johnson, 1979, The Inuence of Fluid Rheology
on the Performance of Traction Drives, in Transactions of ASME, Vol. 101,
July, pp. 266-274.
[19] L. O. Hewko, 1969, EHL Contact Traction and Creep of Lubricated Cylindrical
Rolling Elements at Very High Surface Speeds, in Transactions of ASLE, 12.
[20] B. J. Hamrock and D. Dowson, 1976, Isothermal Elastohydrodynamic Lubri-
cation of Point Contacts Part1-Theoretical Formulation, in Transactions of
the ASME, April.
[21] L. Houpert, 1985, New Results of Traction Force Calculations in Elastohydro-
dynamic Contacts, in Transactions of the ASME , Vol. 107, April.
[22] C. R. Evans and K. L. Johnson, 1986, The rheological properties of elasto-
hydrodynamic lubricants, in Proc. of Mechanical Engineers, 200, c5, pp. 303-
324.
[23] M. O. A. Mokhtar et al., 1987, Elastohydrodynamic Behavior of Elliptical
Contacts Under Pure Rolling Situations, in Transactions of the ASME, Vol.
109, Oct.
[24] W. Hirst and J. W. Richmond, 1988, Traction in Elastohydrodynamic Con-
tacts, in IMechE, Vol. 202, No. C2.
[25] K. T. Ramesh, 1989, On the Rheology of a Traction Fluid, in Transactions
of the ASME, Vol. 111, Oct.
[26] L. Chang et al., 1989, An Ecient, Robust, Multi-Level Computational Al-
gorithm for Elasto-hydrodynamic Lubrication, in Transactions of the ASME,
Vol. 111, April.
128
[27] F. Sadeghi and P. C. Sui, 1990, Thermal Elastohydrodynamic Lubrication of
Rolling/Sliding Contacts, in Transactions of the ASME, Vol. 112, April.
[28] S. Natsumeda, 1992, On the Analysis for the EHL Concentrated Contact, in
Journal of JSLE, Vol. 37, No.12.
[29] M. Muraki and S. Konishi, 1993, Shear Behavior of Low-Viscosity Fluids in
EHL Contacts, in Journal of JSLE, Vol. 38, No. 8.
[30] H. Machida and S. Aihara, 1991, State of the Art of the Traction Drive CVT
Applied to Automobiles, in Vehicle Tribology, Elsevier Science Publishers B.V.
[31] Y. Arakawa et al., 1999, Development and Testing of CVT Fluid for Nissan
Troidal CVT, in Proc. of Inter. Congress on Continuously Variable Power
Transmissions CVT 99, pp. 213-217, Eindhoven, Sept. 16-17, 1999.
[32] Thomas G, Fellows and Christopher J. Greenwood, 1991, The Design and
Development of an Experimental Traction Drive CVT for a 2.0 Liter FWD
Passenger Car, in Transactions of SAE, No. 910408.
[33] P. W. R. Stubbs, 1980, The Development of a Perbury Traction Transmission
for Motor Car Applications, in Journal of Mechanical Design, ASME.
[34] Lubomyr O. Hewko, 1986, Automotive Traction Drive CVTs-An Overview,
in Proc. of Passenger Car Meeting and Exposition, Sep. 22-26, Dearborn, Michi-
gan.
[35] Masaki Nakano and Toshifumi Hibi, 1991, Large Capacity and High Eciency
Toroidal Traction CVT, in Proc. of JSME International Conference on Motion
and Powertransmissions, pp. 965-970, Nov. 23-26, Hiroshima, Japan.
[36] Hirohisa Tanaka and Tatsuhiko Goi, 1999, Traction Drive of a High Speed
Double Cavity Half-Toroidal CVT, in Proc. of Inter. Congress on Continuously
129
Variable Power Transmissions CVT 99, pp. 85-90, Eindhoven, Sept. 16-17,
1999.
[37] Haruyoshi Kumura et al., 1999, Development of a Dual-Cavity Half-Toroidal
CVT, in Proc. of Inter. Congress on Continuously Variable Power Transmis-
sions CVT 99, pp. 65-70, Eindhoven, Sept. 16-17, 1999.
[38] Hisashi Machida, 1999, Traction Drive CVT up to date, in Proc. of Inter.
Congress on Continuously Variable Power Transmissions CVT 99, pp. 71-76,
Eindhoven, Sept. 16-17, 1999.
[39] Hirohisa Tanaka and Masatoshi Eguchi, 1991, Speed Ratio Control of a Half-
Toroidal Traction Drive CVT, in Journal of JSME, pp. 276-279, Vol.57, No.
91-0375B.
[40] Thomas G, Fellows and Christopher J. Greenwood, 1991, The Design and
Development of an Experimental Traction Drive CVT for a 2.0 Liter FWD
Passenger Car, in Transactions of SAE, No. 910408.
[41] H. Tanaka and T. Ishihara, 1984, Electro-Hydraulic Digital Control of Cone-
Roller Toroidal Traction Drive Automatic Power Transmission, in Journal of
Dynamic Systems, Measurement, and Control , pp. 305-310, Vol. 106.
[42] Hirohisa Tanaka and Hisashi Machida, 1991, Stability of a Speed Ratio Control
Servomechanism for a Half-Toroidal Traction Drive CVT in Proc. of JSME In-
ternational Conference on Motion and Powertransmissions, pp. 971-976, Nov.
23-26, Hiroshima, Japan.
[43] Hisashi Machida and Nobuhide Kurachi, 1990, Study of Half Toroidal Contin-
uously Variable Transmission (1st Report, Analysis of Contact Force by means
of Loading Cam), in Journal of JSME, (C), Vol. 56, No. 525, May.
130
[44] Hisashi Machida and Nobuhide Kurachi, 1990, Development of Half Toroidal
CVT -Part 1. Some Examples of Vehicle Test-, in Transactions of JSAE, No.
45.
[45] Takahashi, S., 1998, Fundamental study of low fuel consumption control
scheme based on combination of direct fuel injection engine and continuously
variable transmission, in Proc. of the 37th IEEE Conf. on Decision and Con-
trol , pp. 1522-1529, Tampa, Florida.
[46] K.Sawamura et al., 1996, Development of an Integrated Power Train Con-
trol System with an Electronically Controlled Throttle, in Proc. of Society of
Automotive Engineering of Japan.
[47] L. R. Oliver and D. D. Henderson, 1972, Torque Sensing Variable Speed V-belt
Drive, in Transactions of SAE, No. 720708.
[48] F. S. Jamzadeh and A. A. Frank, 1982, Optimal Control for Maximum Mileage
of a Flywheel Energy-Storage Vehicle, in Transactions of SAE, No. 820747.
[49] C. Chan et al., 1984, System Design and Control Considerations of Automotive
Continuously Variable Transmissions, in Transactions of SAE, No. 8450048.
[50] Ilya Kolmanovsky, Jing Sun, and Leyi Wang, 1999, Coordinated Control of
Lean Burn Gasoline Engines with Continuously Variable Transmissions, in
Proc. of the American Control Conf., pp. 2673-2677, San Diego, California,
June.
[51] S. Liu and B. Paden, 1997, A survey of todays CVT control, in Proc. of the
36th CDC Conference, pp. 4738-4743, San Diego.
[52] Lino Guzzella and Andreas Michael Schmid, 1995, Feedback Linearization of
Spark-Ignition Engines with Continuously Variable Transmissions, in IEEE
131
Transactions on Control Systems Technology, pp. 54-60, Vol. 3, No. 1, March.
[53] Jungyun Kim, Yeongil Park, and Joukou Mitsusida, 1998, S-CVT: the New
Type CVT with Sphere, in Proc. Autumnal Conf. the Korean Society of Pre-
cision Engneers, pp. 815-818, Nov, Seoul, Korea.
[54] Moore, C. A., Peshikin, M. A., and Colgate, J. E., 1999, Design of a 3R Cobot
Using Continuously Variable Transmission, in Proc. of IEEE International
Conference on Robotics and Automation, Detroit, Michigan, May.
[55] Serdalen O. J., Nakamura Y., and Chung W. J., 1994, Design of a nonholo-
nomic Manipulator, in Proc. of IEEE International Conference on Robotics
and Automation, San Diego, CA.
[56] Daniel J. Dawe and Charles B. Lohr, 1993, A High Ratio Multi-Moded Vehicle
Transmission Utilizing a Traction Toroidal Continuously Variable Drive and
Traction Planetary, in Transactions of SAE, No. 932997.
[57] G. Lundberg and A. Palmgren, 1947, Dynamic Capacity of Rolling Bearings,
in Acta Polytechnica, Mechanical Engineering Series, 1, R. S. A. E. E., No. 3.
[58] H. Machida and H. Tanaka, 1991, Oil Film and Surface Damage in Traction
Drive for Automobiles, in Proc. of JSME International Conf. on Motion and
Powertransmissions, pp. 959-964, Nov. 23-26, Hiroshima, Japan.
[59] J. J. Coy, D. A. Rohn, and S. H. Loewenthal, 1981, Constrained Fatigue Life
Optimization of a Nasvytis Multiroller Traction Drive, in Jornal of Mechanical
Design, pp.423-429, Vol. 103, April.
[60] S. H. Loewenthal et al., 1981, Evaluation of a High Performance Fixed-Ratio
Traction Drive, in Transactions of the ASME, Vol. 103.
132
[61] Bor-Tsuen Wang and Robert H.Fries, 1989, Determination of Creep Force,
Moment, and Work Distribution in Rolling Contact With Slip, in Transactions
of the ASME, Vol. 111, Oct.
[62] H. Tanaka et al., 1989, Spin Moment of a Thrust Ball Bearing in Traction
Fluid, in Journal of JSME, 89-0148 B.
[63] E. Rabinowicz, 1965, Friction and Wear of Materials, John Wiley and Sons.
[64] J. Courtney and E. Eisner, 1957, The Eect of a Tangential Force on the
Contact of Metallic Bodies, Proc. of the Royal Society, Vol. A238, pp. 529-
550.
[65] P. R. Dahl, 1977, Measurement of Solid Friction Parameters of Ball Bearing,
Proc. of the 6th Annual Symposium on Incremental Motion, Control Systems
and Devices, pp.49-60, University of Illinois.
[66] N. E. Leonard and P. S. Krishnaprasad, 1992, Adaptive Friction Compensation
for Bi-directional Low-velocity Position Tracking, Proc. of the 31th Conf. on
Decision and Control , pp. 267-273, Tucson, Arisona, Dec. 1992.
[67] Xiaoming Hu, 1994, On Control of Servo Systems Aected by Friction Forces,
Proc. of the 33rd Conf. on Decision and Control , pp. 47-473, Lake Buena Vista,
FL, Dec. 1994.
[68] Chinwon Lee et al., 1999, Comparison of Friction model on the variable DOF
system, Proc. of Spring Conf. the Korean Society of Mechanical Engineers,
Vol. 3, pp. 134-140, Taegu, Korea.
[69] J. W. Gilbert and G. C. Winston, 1974, Adaptive Compensation for an Optical
Tracking Telescope, Automatica, Vol. 10, pp. 125-131.
133
[70] C. Walrath, 1984, Adaptive Bearing Friction Compensation based on Recent
Knowledge of Dynamic Friction, Automatica, Vol. 20, No. 6, pp. 717-727.
[71] J. Craig, 1988, Adaptive Control of Mechanical Manipulators , Addison-Wesley.
[72] C. Canudas, K. Astrom, and K. Braun, 1986, Adaptive Friction Compensation
in DC Motor Drives, in Proc. of the Inter. Conf. on Robotics and Automation,
pp. 1556-1561, April, San Francisco, CA.
[73] C. Canudas de Wit et al., 1995, A New Model for Control of Systems with
Friction, in Transactions of Automatic Control , Vol. 40, pp. 419-425, March.
[74] P. Vedagarbha, D. M. Dawson, and M. Feemster, 1999, Tracking Control of
Mechanical Systems in the Presence of Nonlinear Dynamic Friction Eects, in
Transactions on Control Systems Technology, pp. 446-456, Vol. 7, No. 4, July.
[75] Hirohisa Tanaka, 1986, Power Transmission of a Cone Roller Toroidal Traction
Drive (1st report, Speed and Torque Transmission Eciencies), in Journal of
Japan Society of Mechanical Engineers , No. 86-1182A.
[76] S. Timoshenko and J. N. Goodier, 1951, Theory of Elasticity, 2nd Ed., McGraw-
Hill Book Co.
[77] Sunmo Chung and D. C. Han, 1996, Standards of Mechanical Design,
Dongmyung-Sa, Korea.
[78] A. Isidori, 1995, Nonlinear Control Systems, 3rd Ed., Springer-Verlag.
[79] A. Isidori, Maria D. Di Benedetto, 1996, Feedback Linearization of Nonlinear
Systems, in The Control Handbook, Vol. 2, pp. 909-917, CRC and IEEE Press.
[80] Hassan K. Khalil, 1996, Nonlinear Systems, 2nd Ed., Prentice Hall.
[81] Jean-Jacques E. Slotine and Weiping Li, 1991, Applied Nonlinear Control,
Prenice Hall.
134
[82] Bahram Shahian and M. Hassul, 1993, Control System Design using MATLAB,
Prentice Hall.
[83] Leonard Meirovitch, 1986, Elements of Vibration Analysis, Chapter 2, 2nd Ed.
McGraw-Hill Co.
[84] Werner Leonhard, 1996, Control of Electrical Drives, 2nd Ed., Springer.
[85] J. G. Kassankian, M. F. Schlecht, and G. C. Verghese, 1991, Principles of Power
Electronics, Cambridge, MA, Addison-Wesley, Chapter 1-8.
[86] Tadashi Egami, Hideaki Morita, and Takeshi Tsuchiya, 1990, Eciency op-
timized Model Reference Adaptive Control System for a dc Motor, in IEEE
Transactions on Industrial Electronics, Vol. 37, No. 1, February.
[87] Parviz Famouri and Wils L. Cooley, 1994, Design of DC Traction Motor Drives
for High Eciency Under Accelerating Conditions, in IEEE Transactions on
Industry Applications, Vol. 30, No. 4, July/August.
[88] J. Borenstein, H. R. Everett, and L. Feng, 1996, Where am I? Sensors and
Methods for Mobile Robot Positioning, The University of Michigan, April.
[89] J. L. Jones and A. M. Flynn, 1993, Mobile Robots: Inspiration to Implementa-
tion, Massachusetts, A K Peters.
[90] Jungyun Kim, F. C. Park, and Y. Park, 2000, Output Position Changing De-
vice for Continuously Variable Transmission, granted patent No. KR 10-2000-
0043861, pending by the Oce of the Patent Administration of Korea.
[91] Robert C. Weast, and Melvin J. Astle, William H. Beyer, 1987, CRC Handbook
of Chemistry and Physics, pp. F16-F18, 68
th
Ed., Boca Raton, Florida, CRC
Press, Inc.
135
[92] H. S. Jo, W. S. Lim, J. M. Lee, Jungyun Kim, 1996, Dynamic Performance
Estimation and Optimization for the Power Transmission of a Heavy Duty Vehi-
cle, in Transactions of Korea Society of Automotive Engineers, No. 96370006.
[93] Jungyun Kim, et al., 1996, Innitely Variable Speed Transmission, patent No.
96907771.8-2306, granted by the European Patent Administration.
[94] Jungyun Kim, Y. D. Choi, I. W. Yeo, 1997, Development of Simulation Pro-
gram for Launch Performance of CVT equipped Passenger Car, in Technical
Reviews 97 of Daewoo Motor Company, pp. 227-232, Sep., Daewoo Motor
Company.
[95] Jungyun Kim, Jang Moo Lee, In Wug Yeo, 1999, A Development of the Simu-
lation Program for Launching Performance of a Passenger Car equipped Contin-
uously Variable Transmission, in Transactions of Korea Society of Automotive
Engineers, No. 99370231, pp. 157-166, Vol. 7, No. 7.
[96] Jungyun Kim, Han Jun Yeom, F. C. Park, 1999, MOSTS: A Mobile Robot
with a Spherical Continuously Variable Transmission, in Proceedings of the
1999 IEEE/RSJ International Conference on Intelligent Robots and Systems,
pp. 1751-1756, Vol. 3, Oct.17-21, Kyoungju, Korea.
[97] Jungyun Kim, Hyoungjun Cho, Jangmoo Lee, Won Sik Lim, 1999, A Study of
the Launching Characteristics for a passenger car having a CVT with Dierent
Torsional Couplings, in Proceedings of the 1999 Autumnal Conference of Korea
Society of Automotive Engineers, No. 99380247, pp. 770-775, Vol. 2, Nov. 26-27.
[98] Jungyun Kim, Hanjun Yeom, F. C. Park, Y. I. Park, Munsang Kim, 2000, On
the Energy Eciency of CVT-Based Mobile Robots, in Proceedings of IEEE
International Conference on Robotics and Automation, pp. 1539-1544, Vol. 1,
April 24-28, San Francisco, CA, U.S.A.
136
[99] Jungyun Kim, Yeongil Park, F. C. Park, Jangmoo Lee, 2000, On the De-
velopment and Application of the Spherical CVT, in Proceedings of the 2000
Spring Conference of Korean Society of Mechanical Engineers, No. 00S118, pp.
690-695, Vol. A, April 21-22, Ulsan, Korea.
[100] Hyoungjun Cho, Jungyun Kim, Soonil Jeon, Jangmoo Lee, Jaeyul Kim,
Yeongil Park, 2000, Development of a Hybrid Electric Vehicle equipped with
Semi-Spherical CVT, in Proceedings of the 2000 Spring Conference of Korea
Society of Automotive Engineers, No. 2000-03-2054, pp. 349-354, Vol. I, April
28-29, Seoul, Korea.
[101] Jungyun Kim, et al., 2000, Apparatus for Multi-mode Continuously Variable
Transmission, granted patent No. KR 10-2000-0054436, pending by the Oce
of the Patent Administration of Korea.
[102] Jungyun Kim, F. C. Park, Yeongil Park, Joukou Mitsusida, 2000, Design
and Analysis of a Spherical Continuously Variable Transmission, Accepted in
Journal of Mechanical Design, ASME.
137