International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

Sensitivity Analysis Study of CVT Parameters Using Mathematical Model

NIKHIL SINGH1, SURAJ S1, BHARANI S ANAND1, SUPREETH A GOWDA1
1Students, Dept. of Mechanical Engineering, JSS Academy of Technical Education, Bengaluru, Karnataka, India

---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – The paper presented here discusses one of the
most valuable invention in the field of automobile engineering,
the Continuously Variable Transmission or the CVT.
Although the invention of CVT dates back to 1490s there is a
lot more to learn about this technology. This paper will
present the current problem of manufacturing and assembly of
the CVT setup of a two wheeler, and also discuss how a small
variation in the CVT parameter such as length of V-belt, mass
of the rollers, spacer length, angle between the pulley, center
distance between the pulley, even within the tolerance range
can cause drastic changes in the performance of the vehicle. Fig -1: General Layout of the CVT.
This paper will also study how and what kind of different
combinations of above mentioned CVT parameters( keeping COMPONENTS OF A CVT SYSTEM
one or the other parameters constants) will cause different  A V type rubber belt
effects on vehicle performance parameters such as fuel  A variable-input "driving" pulley
efficiency, acceleration time , reliability etc. All of this will be  An output "driven" pulley
studied using a mathematical model, such that various outputs
can be generated for various inputs at least possible time, and
plotting these results as a graphical representation.

Key Words: CVT performance, Mathematical Model,

Microsoft Excel, Roller Mass, Spacer Length, Pulley, V-
Belts.

1. INTRODUCTION

Continuously Variable Transmission or simply CVT is the

most common power transmission system in the field of
automobile. In general, there are two pulleys that are divided Fig -2: Components of CVT
perpendicular to their axis of rotation, and with a V-belt
running in between them. The gear ratio can be The V- belt and the variable-diameter pulleys are the main
automatically changed by moving the two sections of one part of the CVT system. Each pulley is made of two 23-degree
pulley closer together and the two sections of the other cones facing each other. A belt drives in the seams between
pulley little bit apart. Due to the V-shaped cross section of of the two cones of the pulleys. V-belts are preferred if the
the belt, this causes the belt to ride higher on one pulley and belt is made of rubber. When the two cones of the pulley are
lower on the other. Because of this there are changes in the far apart (i.e, the diameter increases), the belt drives lower
effective diameters of the pulleys, which changes the overall in the seams, and the radius of the belt loop going around the
gear ratio. The distance between the two pulleys cannot pulley gets smaller. When the cones are close together (i.e.,
change, and neither does the length of the V- belt changes , the diameter decreases), the belt drives higher in the seams,
so changing the gear ratio means both pulleys must undergo and the radius of the belt loop going around the pulley gets
adjustment (one can bigger, the other smaller, or both can be larger. CVTs may use hydraulic pressure, centrifugal force or
of same size) simultaneously to maintain the proper amount spring tension to create the force necessary to adjust the
of tension on the belt. The V-belt needs has to be very stiff in pulley halves. Variable-diameter pulleys must always come
the pulley's axial direction in order to make only small radial in pairs. One of the pulleys, known as the drive pulley (or
movements while moving in and out of the pulleys. driving pulley), is connected to the crankshaft of the engine.
The driving pulley is also called the input pulley because it's
where the energy from the engine

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

neglected. For the primary (driving) actuator, only the radial

static friction is assumed. Thus, single contact area is
considered. The total axial force for the primary actuator is
expressed by equation (3) and for the secondary actuator the
two contact areas are assumed.
In the passive contact area (at the entrance), the belt tension
is constant. In this area the belt element forces distribution is
familiar with the belt element forces distribution on the
primary actuator. The kinetic force in radial and
circumferential direction is considered in the active contact
area. The total axial force of the secondary actuator is
expressed by equation (4).
Fig -3: Working Principle of CVT
1.1.2 The Cammalleri model
enters the transmission. The second pulley is called the
driven pulley because the first pulley is turning it. As an
The main difference between the Kim-Kim described
output pulley, the driven pulley transfers energy to the
models and the Cammalleri model [1] lies in the
driveshaft. When one pulley increases its radius, the other
consideration of the belt flexibility and the changes of the
decreases its radius to keep the belt tight. As the two pulleys
slip angle along the pulley wrap angle.
change their radii relative to one another, they create an
infinite number of gear ratios -- from low to high and
everything in between. When the pitch radius is small on the 1.2 OBJECTIVES
driving pulley and large on the driven pulley, the rotational
speed of the driven pulley decreases resulting in a lower 1. Create a Mathematical Model of the sensitivity analysis
gear ratio. When the pitch radius is large on the driving of the CVT parameter.
pulley and small on the driven pulley, then the rotational 2. Verification of the Mathematical Model with the practical
speed of the driven pulley increases, resulting in a higher values.
gear ratio. Thus, in theory, a CVT has an infinite number of 3. Plot graphs for different parameters and analysis of how
"gears" that it can run through at any time, at any engine or the changes in the CVT parameters effects the vehicle
vehicle speed. The simplicity and steeples nature of CVTs performance.
make them an ideal transmission for a variety of machines
and devices, not just automobiles. CVTs have been used for 2. NOMENCLATURES
years in power tools and drill presses. They've also been
used in a variety of vehicles, including tractors, snowmobiles 1. FR Axial Force Produced by Rollers
and motor scooters. In all of these applications, the 2. M Mass of Rollers
transmissions have relied on high-density rubber belts, 3. yg Distance of Roller Mass center G from pulley axis
which can slip and stretch, thereby reducing their efficiency. 4. β Angle Formed by the Pulley Axis and the Tangent
The distance between the centers of the pulleys to where the To the Curved Profile at Contact Point
belt makes contact in the seams is known as the pitch radius. 5. ϒ Angle of the Back Plate with Roller
When the pulleys are far apart, the belt rides lower and the 6. 𝜑sx Angle of Sliding Friction between Roller and Ramp
pitch radius decreases. When the pulleys are close together, 7. 𝜑dx Angle of Friction of the Guide
the belt rides higher and the pitch radius increases. 8. FO Pre Load on the Secondary Actuator Spring
9. K Stiffness of the spring
1.1 LITERATURE REVIEW 10. D Mean Diameter of Spring
11. d Mean Diameter of the Helical Guide
12. 𝜁 Angle of Helical Guide Slope
1.1.1 The Kim Kwangwon & Hyunsoo Kim Model
13. V Poisson’s Ratio
14. Ca Torque at Primary Actuator
According to this method of modelling, the meandering
13. Cb Torque at Secondary Actuator
radius on both the actuators must undergo the changes
14. Fr Axial Force of the Driver (Primary) Actuator
because of the belt force differentiation. The experimental
15. Fn Axial Force of the Driven (Secondary) Actuator
research results, described in the work [4], indicate that such
16. T1 Tension on Tight Side of the Belt
belt behaviour occurs only for the driven wheel. For the
17. T2 Tension on Slack Side of the Belt
driving wheel, the meandering radius remains same on the
18. θa Active Wrap Angle
complete wrap angle. This behaviour can be explained with
19. θ Length of Arc of Contact
the help of self-locking phenomenon occurrence.
20. α V-Belt Wedge Angle
As a result, it is presumed that the force exerted by the belt
21. µ Co-efficient of Friction between Belt and Actuator
is constant. Thus, the circumferential friction component is
22. Ra Active Radius of Driver (Primary) Actuator
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

23. Rb Active Radius of Driven (Secondary) Actuator This actuator consists of number of centrifugal mass rollers
which are placed on a curved profile called ramp, so as the
3. METHODOLOGY speed of the engine increases which indeed increases the
speed of the primary actuators and vice versa, there is a
The tests was carried out on a Dynamometer test rig. centrifugal force which is acting on these rollers which
The test gives us the following outputs creates an axial force and acts on the back plate, thereby
i. The engine RPM with respect to the vehicle speeds at pushing the back plate forward or backward depending on
different loading conditions. the speed of the actuators and thereby varying the radius of
ii. The wheel force. the belt. The free body diagram of the roller is given in the
Using these values as input we determine the various figure below.
parameters that affects the CVT performance.

I. The Primary Actuator.

The primary actuator or the primary pulley is directly

connected to the crank shaft of the engine and thereby it is
assumed that the speed of the primary actuator (na) is equal
to the engine speed in rotations per minute (RPM).

Fig -5: Free Body Diagram

The figure 2 shows FBD for the case upshift i.e, increase in
the speed ratio. Where Fcen is the centrifugal force acting on
the rollers. Imposing radial, axial, and rotational equilibrium,
it is possible to write:-

____(1)

____(2)

____(3)

As Fcen=Mn2ayg and FR= Fdx sin(β− 𝜑sx), the equations 1

and 2 gives the axial forced produced by the centrifugal
rollers and is given by the below equation

____(4)
From equations 2 and 3 the roller motion is given by

Fig -4: Primary Actuator

____(5)

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 3
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

It is possible to come to a conclusion that,

for ,

[𝜑dx from equation (5)] and that the roller rolls on the back
plate and slips on the left ramp.
While, for

[𝜑sx from equation (5)] and the roller rolls on the left ramp
and slips on the back plate.
(Where 𝜑a is the angle of sliding friction between the
roller and the ramp).
And is given by the equation:- Fig -6: V-Belt Geometry

Using the Speed Ratio-Torque and Thrust Relationship

The axial force that is developed by the belt on the primary
In the case of downshift, the frictional forces reverse their actuator can be given by the equation below
angles 𝜑dx and 𝜑sx and the directions change their sign in the
above equations, but all the results on the roller motion
remain the same.

Yg is determined with the help of CAD drawing of the primary

actuator by finding 3 different roller position from the ____(6)
rotational axis to the centre of the roller and also the angle
tangent to the ramp (α). Then using polynomial equation the Also the axial force that is developed on the secondary
roller position along with tangent angle can be found at any actuator due to the V-belt is given by the equation
instant.

___(7)
Where,

____(8)

Using the value of Fr obtained in equation 4 and substituting

Chart -1: Determining the position and angle of roller it in the equation 6 to obtain the value of T1. Then by using
the belt equation given below
II. V-Belt

The V-belt is the linking member between the primary and

secondary actuator. It is known that the belt is elastic in
nature and due to this property of the belt, it penetrates into ___(9)
the side walls of the actuators and due to which there is a Finding the value of T2 from the value obtained by equation
radial component of frictional force that comes into action 6. Then by substuting the value of T1 and T2 obtained from
and this plays an important role in the V-belt mechanics. equations 6 and 8 respectively into equation 7 to find the
value of the axial force developed by the V-belt on the
secondary actuator.

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 4
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

III. Secondary Actuator

The torque that is generated at the secondary actuator (Cb) is
The axial force that is needed to assure the connection obtained by substuting the value of Fn from equation 7 in
between the V-belt and the secondary actuator is generally equation 10. As the output of the dynamometer test is
produced by a helical spring compressed in between the measured in terms of the wheel force, the torque obtained
half-secondary movable pulley and a fixed contrast wall. In from the equation 10 need to be converted into wheel force,
addition , in motorcycle actuators , an extra force which is for which torque is multiplied by the reduction gear ratio
proportional to the output torque is generated by the slope and divided by the dynamic rolling radius of the wheel. Using
of the sliding guides of a pre-defined constant angle 𝜁 as all the formulas above an excel table is formulated in order
shown in the figure below. to find various parameters at different speeds of the vehicle.

4. RESULTS

At first it is important to verify that the mathematical model

hold true or not, for that a dyno test was conducted on the
standard vehicle and results were compared with the
mathematical model and a graph representing this
comparison is shown below:-

Chart -2: Verification of the Mathematical Model

As it can be seen the blue line represents the mathematical

model values and orange line represents experimental
values and both of them are comparable, for speeds upto
15Kmph the values divert a lot due to slippage of the belts
on the actuators. Now we can compare various CVT
Fig -7: Secondary Actuator parameters using this mathematical model.

The equation for axial force equilibrium of the movable half

actuator is given by

____(10)
Depending on the constraints on the spring, the half-actuator
movement may cause some torsional effect on the spring. In
such situation the reaction torque produces an extra axial
thrust, which is the function of speed ratio. All together, the
equation of the actuator axial force remains the same
provided that the spring stiffness is given by

____(11) Chart -3: Varying the mass of the rollers

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 5
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

Similarly all other parameters such as Length of the V-belt, [8] Dolan, J. P., and W. S. Worley. "Closed-form
Pulley wedge angle, center distance between the pulley, approximations to the solution of V-belt force and slip
spacer length and spring stiffness can varied in the equations." Journal of Mechanisms, Transmissions, and
Automation in Design 107.2 (1985): 292-300.
mathematical model and can be studied for each and every
[9] Kong, Lingyuan, and Robert G. Parker. "Mechanics and
combination of the parameter to obtain the best results in sliding friction in belt drives with pulley
terms of power or torque transfer depending on the users grooves." Journal of Mechanical Design 128.2 (2006):
need without conducting the actual test by varying each 494-502.
parameters. [10] Julio, G., and J-S. Plante. "An experimentally-validated
model of rubber-belt CVT mechanics." Mechanism and
5. CONCLUSIONS Machine Theory 46.8 (2011): 1037-1053.

After carrying out various test by varying the combination of BIOGRAPHIES

the CVT parameters the following conclusion are drawn:-
1. Mass of the rollers: - The wheel force increases by
decreasing the mass of the rollers and vice versa. NIKHIL SINGH, Graduated in
2. Length of Belt: - The wheel force decreases with increase B.E Mechanical Engineering, JSS
in the length of V-Belt, in other words wheel force is
Academy of Technical
inversely proportional to the length of belt.
3. The Wheel Force increases with increase in the wedge
Education affiliated to VTU
angle of the pulley, center distance of the pulley and Belagavi, 2018 Batch.
spring stiffness.

ACKNOWLEDGEMENT

We would like to thank TVS Motor Company, Hosur for SURAJ S, Mechanical
letting us carry out our research at their state of the art Engineering student at JSS
Research and Development center, and guiding us in using Academy of Technical
advanced equipments. Education affiliated to VTU
Belagavi.
REFERENCES

[1] Cammalleri, M. A. R. C. O. "A new approach to the design

of a speed-torque-controlled rubber V-belt
variator." Proceedings of the Institution of Mechanical
Engineers, Part D: Journal of Automobile
Engineering 219.12 (2005): 1413-1427. BHARANI S ANAND, Graduated
[2] Cammalleri, M., A. Conti, and F. Sorge. "Experimental in B.E Mechanical Engineering,
results for a belt variator in transient conditions." XVIII JSS Academy of Technical
Congresso AIMETA. 2007.
Education affiliated to VTU
[3] Grzegożek, Witold, Krzysztof Dobaj, and Adam Kot.
"Experimental verification and comparison of the Belagavi, 2018 Batch.
rubber V-belt continuously variable transmission
models." IOP Conference Series: Materials Science and
Engineering. Vol. 148. No. 1. IOP Publishing, 2016.
[4] Kim, Kwangwon, and Hyunsoo Kim. "Axial forces of a V-
belt CVT." KSME Journal 3.1 (1989): 56-61.
[5] Kim, Hyunsuk, et al. "Steady state and transient
characteristics of a rubber belt CVT with mechanical SUPREETH A GOWDA,
actuatiors." KSME international journal 16.5 (2002): Mechanical Engineering student
639-646. at JSS Academy of Technical
[6] Oliver, Larry R., et al. Design equations for a speed and Education affiliated to VTU
torque controlled variable ratio V-belt transmission. No.
730003. SAE Technical Paper, 1973. Belagavi.
[7] Zhu, C. C., et al. "Experimental research on the effect of
structural parameters on the governing characteristics
of a pulley-drive, continuously variable
transmission." Proceedings of the Institution of
Mechanical Engineers, Part D: Journal of Automobile
Engineering 224.6 (2010): 775-784.

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 6