Parametric Study of A Dog Clutch Used in A Transfer Case For Trucks
Parametric Study of A Dog Clutch Used in A Transfer Case For Trucks
Parametric Study of A Dog Clutch Used in A Transfer Case For Trucks
Växjö, 2013-05-31
Mechanical Engineering
Authors:
Amanuel Mehari, Fredrik Eriksson and Linu Kuttikal Joseph
Supervisor, LNU: Andreas Linderholt
Examiner: Andreas Linderholt
Semester: Spring 2013, 15 credits
Course code: 4MT01E
Organization /Organisation Författare/Authors
Linnaeus University, Växjö, Sweden Amanuel Mehari, Fredrik Eriksson and
School of Engineering Linu Kuttikal Joseph
Summary
This thesis work consists of an investigation of design changes of a dog clutch. The dog clutch
studied is being used in the transfer case of trucks having four wheel drives.
Normally the trucks with four wheel drive option will be running in rear wheel drives and the
front wheels will be rotating freely. In extreme tough driving conditions, the risk for slipping of
the rear wheels or getting stopped in mud is high. By using four wheel drives, which makes all the
wheels drive the vehicle, the traction and the maneuverability of the vehicle will be high. When
the driver tries to engage the four wheel drive option, due to the difference in relative rotational
speed of the dog clutch parts, there is a risk for slipping off or bouncing back of the dog clutch.
Therefore it is inevitable to improve the dog clutch design to increase the performance while
engaging the four wheel drives when the vehicle is running. Here, the gear geometry and a few
parameters of the dog clutch is modified. The efficiency of the new design found satisfactory
when simulated in MSC ADAMS.
Keywords
Dog Clutch, Four wheel drive, MSC ADAMS, Multibody dynamics, Latin Hypercube,
SolidWorks.
II
Abstract
Normally the trucks with four wheel drive option will be running in rear wheel drives and the
front wheels will be rotating freely. In extreme tough driving conditions, the risk for getting
stopped or slipping the rear wheels in mud is high. When the driver tries to engage the four wheel
drive option and due to the difference in relative rotational speed of the dog clutch parts, there is a
risk for slipping off or bouncing back of the dog clutch.
After studying the importance of gear geometry and a few parameters, the team ended up with a
new design and the performance of the design found satisfactory when simulated in MSC
ADAMS.
III
Acknowledgement
This thesis work would not have been possible without the guidance and the help of
several individuals who in one way or another contributed and extended their valuable
assistance in the preparation and completion of this study.
First and foremost, our utmost gratitude to our lecturer Andreas Linderholt (School of
Engineering, Linnaeus University, Växjö, Sweden) for supervising this thesis work.
Second, Mr. Hans Hansson and Ms. Karolina Årdh, technical engineers working for
SwePart Transmission AB, Liatorp, Sweden for giving the opportunity to conduct this
thesis work and explaining the concern moreover to provide us with the necessary
information.
IV
Table of contents
1. Introduction......................................................................................................1
1.1 Background ...................................................................................................................................... 1
1.2 Problem description ....................................................................................................................... 2
1.3 Purpose and goals ........................................................................................................................... 3
1.4 Limitations ........................................................................................................................................ 4
2. Theory.............................................................................................................. 5
2.1 MSC ADAMS2, the multi body simulation software .......................................................... 5
2.2 SolidWorks ....................................................................................................................................... 6
2.3 System engineering......................................................................................................................... 7
2.4 Product development ..................................................................................................................... 8
2.5 Structural Dynamics ...................................................................................................................... 9
2.6 Global and Local coordinate system ...................................................................................... 10
2.6.1 Global Coordinate system9 .................................................................................................. 10
2.6.2 Local coordinate system9 ..................................................................................................... 11
2.7 Latin Hypercube ........................................................................................................................... 12
4. Method ............................................................................................................ 14
4.1 The main methods ........................................................................................................................ 14
4.4 Drawings in SolidWorks ............................................................................................................ 19
4.5 Analysis in MSC ADAMS .......................................................................................................... 19
5. Result ............................................................................................................. 20
5.1 Simulation results ...................................................................................................................... 20
5.2 Simulation procedure ............................................................................................................... 21
5.3 Result summary ......................................................................................................................... 22
7. References...................................................................................................... 26
8. Appendix........................................................................................................ 28
V
1. Introduction
1.1 Background
In the year of 1945, the two brothers Bertil Bengtsson and Axel Bengtsson
started the engineering workshop Mekan AB, at Hultaberg in the southern part of
Sweden. The business was to manufacture screw-vices, brass-spindles and
hydraulic equipments. Later in the year 1995 Mekan AB changed its name to
SwePart Transmission AB; an ISO 9001 engineering company which designs
and manufactures most precised gears, shafts, cross joints, gear boxes according
to the customers’ specifications. In 2008 Swepart Transmission AB had a
turnover of around 350 million SEK.
This thesis work consists of the investigation of a transfer case used in trucks
with four wheel drives. Usually the trucks with four wheel drive option will be
running in rear wheel drives and the front wheels will be rotating freely. In
extreme tough driving conditions, the risk for slipping of the rear wheels or
getting stopped in mud is high. By using four wheel drives, which makes all the
wheels drive the vehicle, the traction and the maneuverability of the vehicle will
be high. The four wheel drive is engaged by a shift lever which is positioned near
to the gear lever in the drivers’ cabin, see figure 1. When it is engaged, the front
wheels also have driving torque from the engine through the dog clutch which is
positioned inside the transfer case. The location of the transfer case is between
the transmission and the rear differential, see figure 2.
14
Figure 1. Position of 4x4 shift lever
1
Generally the trucks have rear wheel drives whereas the front wheels are rotating
freely. This can be changed into four wheel drives using a dog clutch which is
positioned inside a unit called a transfer case, see figure 2. The engagement and
disengagement of the dog clutch inside the transfer case is controlled by the
driver from the drivers cabin using a shift lever mechanism, see figure 1. If the
truck is forced to stop or get stuck on a slippery surface with full pay load, it may
be hard to move the vehicle from rest even after engaging the four wheel drives.
Moreover, while trying to engage the four wheel drives without stopping the
truck, at times, the dog clutch will bounce back which often makes metallic noise
without engaging. This is happening because the angular velocities between the
two dog clutch halves are not equal.
1
Figure 2. Location of Transfer case
One solution is to first design a transfer case with a synchronizer and then to use
sensors for the wheels for measuring the rotation continuously. The data from the
sensors calculate the slip of the rear wheels thus engaging the four wheel drives
automatically if required. Since the solution involve high manufacturing cost, our
team was challenged to come up with another solution.
The purpose of this thesis work is to evaluate the importance of the gear
geometry, mass, rotational speed, material stiffness and engagement force in the
dog clutch being used in four wheel drives trucks and to investigate whether the
existing design can be improved or not.
A ten week thesis period is insufficient for a bigger project like the parametric
study of a dog clutch.
The parameters such as friction and engagement acceleration of the dog clutch
have a constants value.
Moreover the testing will be carried out in a later stage, this is outside the scope
of this thesis work.
MSC ADAMS is a software with which engineers can investigate and simulate
mechanical systems regarding dynamics, statics and kinematics.
MSC ADAMS can simulate how a vehicle will perform for different road
conditions. Before using MSC ADAMS, many automobile manufactures
conducted real tests on test tracks to study the performance as a function of
different parameter settings.
SolidWorks has over 1.4 million users in more than 165 000 companies all over
the world.
Pressures
Forces
Temperatures
Accelerations
Contact between components
6
Figure 3. Life cycle approach in system engineering
These steps in the lifecycle approach may be iterated if required and some may
be eliminated.
The philosophy of the lifecycle approach can be useful in any system at all
levels.
The goal of product development7 is to bring something new that satisfy the
customers need and preferably surpasses their need.
1) Idea Generation.
- Making SWOT analysis (Strengths, Weaknesses, Opportunities
and Threats).
- New ideas are created around the new product , some of these
ideas are implemented.
2) Idea Screening.
- Will this satisfy the customers?
- Will this be gainful when manufactured and supplied at the aim
price?
3) Concept Development and Testing.
- Improve both engineering and marketing details.
- How can this product be manufactured most cost effectively?
- What profits will the product incorporate?
4) Business Analysis.
5) Beta Testing and Market Testing.
6) Technical Implementation.
- Supply estimation.
- Requirement publication.
- Announce technical communications such as data sheets.
- Contingencies – what – if planning.
7) Commercialization.
- Introduction the product.
8) New Product Pricing.
- Impact of new product on the entire product portfolio.
- Product cost.
These stages can be iterated if needed and some can even be eliminated.
Dynamics8 deals with the study of bodies using tests and mathematical models.
In the field of engineering, structural dynamics use the principle of virtual
displacements which eliminates the necessity of using interaction forces directly.
Even though this can be used to derive the equations of motion for multi degree
of freedom (MDOF) system, it is much more convenient and simpler to use the
Lagrange’s equation shown in equation (1).
( ̇
) [Equation 1]17
T: Kinetic energy
V: Potential energy
L=T–V [Equation 2]
The equations of motion are used to describe the behaviors of physical systems.
̈ ̇ ( )
M: Mass matrix
C: Damping matrix
K: Stiffness matrix
This will allow the use of scalar quantities such as work, potential energy and
kinetic energy instead of vector quantities such as force and acceleration.
Moreover this can be derived from principle of virtual displacements or extended
Hamilton’s principle.
The global coordinate gives the frame in which the model is based. To filter the
useful information from this environment, it is inevitable to study from the
multiple phased arrays. These arrays denote the location from its own local
coordinate system. In order to arrange this kind of information from each phased
array into a global coordinate, it is necessary to know the position and location of
each array in global coordinate system.
For example in figure 4 shown below, the object is the black sphere. The thick
black axes represent the coordinate axes of the global coordinate system.
However, two phased arrays 1 and 2 in figure 4, represent their own coordinate
system within the global coordinates which is represented by dashed line. While
specifying a global coordinate system, it designates any point as the origin and
the coordinate axes must be orthogonal.
10
11
This method was first described by McKay in 1979. In this type of sampling, a
square grid containing sample positions is a latin square if and only if there is
only one sample in each row and column. For example when sampling X
variables, the range of each variable is divided into Y equally intervals. Then Y
sample points are fixed to satisfy the Latin hypercube requirement which
determines the number of divisions, Y, which should be equal to each variable.
Choose the number of simulation tests required, maximum and minimum values
for each parameter. Finally inserting these values in Matlab gives output for
making the simulations in MSC ADAMS.
12
This thesis work sets out from previous work, here referred to, made at Linnaeus
University, Växjö.
In the first thesis work made by two former students at Linnaeus University, the
dog clutch geometry was analyzed using the FEM code Abaqus. In their thesis12
work, analyses of how the contact pressure acts on the clutch teeth and how that
changed as a result of parameter changes such as chamfer distance, chamfer
angles, teeth angle and number of teeth was made.
In second thesis11 the dog clutch engagement was analyzed with help of MSC
ADAMS instead of Abaqus, including these parameters.
By determining the moment of inertia and stiffness of the rotating drive shafts
they made a model in MSC ADAMS see in figure 6. The dog clutch was
connected with a torsion spring to a cylinder (number 3) with a mass moment of
inertia, finally the dog clutch (number 1 and 2) is engaging through an
engagement force.
This thesis work investigated a new design idea from Swepart Transmission AB.
Here the geometry of the teeth for dog clutch was analyzed and compared to the
existing design.
13
14
Top land: Top land is the (sometimes flat) surface of the top of a gear tooth.
Bottom land: The bottom land is the surface at the bottom of a gear tooth space
adjoining the fillet.
Face width: It is the length of teeth in an axial plane. For double helical, it does
not include the gap.
Face: The Face of a gear tooth is the upper half of the gear tooth contact area.
Flank: The Flank of a gear tooth is the lower half of the gear tooth contact area.
Circular pitch: Distance from one face of a tooth to the corresponding face of an
adjacent tooth on the same gear, measured along the pitch circle.
Circular thickness: Length of arc between the two sides of a gear tooth, on the
specified datum circle.
Chordal pitch: Distance from one face of a tooth to the corresponding face of an
adjacent tooth on the same gear, measured straight from one point to another.
Pitch circle: Circle centered on and perpendicular to the axis, and passing
through the pitch point. A predefined diametric position on the gear where the
circular tooth thickness, pressure angle and helix angles are defined.
Addendum ( a )Radial distance from the pitch surface to the outermost point of
the tooth.
Dedendum ( b ) Radial distance from the depth of the tooth trough to the pitch
surface.
Base circle: In involute gears, where the tooth profile is the involute of the base
circle. The radius of the base circle is somewhat smaller than that of the pitch
circle.
The addendum circle: It coincides with the tops of the teeth of a gear and is
concentric with the standard (reference) pitch circle and radially distant from it
by the amount of the addendum.
The root circle: It coincides with the bottoms of the tooth spaces.
15
Figure 8 shows teeth angles (marked as 9º & 4.5º) and the chamfer distance ‘b’.
Figure 9, shows term chamfer angles (marked as 30º, 60º and 45º) and the
chamfer distances (marked1.5 and 1). Furthermore the term top land (marked
‘A’) and the bottom land (marked ‘B’) are shown in the figure below.
Figure 9. Defining the chamfer angle, distance, top land and bottom land
16
The first risk is that the clutch halves bouncing off each other, see (1) in figure
10.
The second is slipping off each other, see (2) and (3) in figure 10.
The third is engaging the first time without any problem, see (4), (5) and (6) in
figure 10
17
When one end of the dog clutch is fixed mimicking a connection to an infinite
mass (flywheel) or an infinite torsional stiffness the dog clutch see figure 12, is
not engaging, but bouncing.
Figure 12. Before engaging, when connected to larger mass or torsional stiffness.
18
The drawings are made in SolidWorks. The first step is to sketch a 2-D drawing
of the dog clutch. This is converted into a 3-D model by using revolve option in
SolidWorks. Other settings such as fillet, extrude, circular pattern, chamfer and
cut-revolve are used for making the required design. The drawings are imported
into MSC ADAMS by changing the format to parasolid (x. t).
To analyze the real world problems, the mathematical representations are more
or less simplifications. In this case simplifications are made regarding the
geometry, engagement force and material property
19
18
Figure 13. Model YY
Figure 14. Showing the method for studying of the dynamics as a function of parameter values
To achieve the best results, the simulation process should be automated, since
that it is less time consuming and can avoid human error. The process shown in
figure 13, starts with a Latin hypercube sampling using Matlab. After inserting
the lower and upper limits of each parameter in Matlab, combinations of
parameter values are extracted. This thesis work uses ten parameter
combinations for simulations in MSC ADAMS. For controlling the simulations,
the models have been exported from SolidWorks and imported to MSC
ADAMS.
In this thesis, the design is simulated with two different starting positions. The first
starting position is teeth to teeth and the second is teeth to gap, see figure 15 A) and
B) respectively. To find the best result using both starting positions, analysis of the
designs is carried out.
Figure 15. A) Starting position teeth to teeth B) Starting position teeth to gap
The female part, see figure 16, is fixed in the x- y- and z-axis, but is free to rotate
around the x-axis. The male part can move along the x-axis and is free to rotate
with the female part but it is fixed in both y- and z-axis. After one second of
simulation, the male part starts to move along the x-axis due to a prescribed
force. During the engagement the male part bounces off at times or slides over
the female part because of the difference in the angular velocity. During this
process the male part may get a small amount of angular velocity from the
female part. However, it engages before or after a couple of trials.
Figure 16. A picture showing the difference between a female part and a male part
21
The red line represents the starting position of the male dog clutch is, teeth to
teeth, with respect to the female and the blue dashed line in the graph represent
the starting position of the male dog clutch, teeth to gap, with respect to the
female.
After engagement, the distance in y-axis will vary for different models because
of the difference in the designs and parameter combinations.
This thesis work considered four different parameters such as mass [kg], rotation
[degree/second], material stiffness [newton-meter/degree] and the engagement
force [newton], see table 1 in next page.
22
Parameters
Simulation test
Mass Rotation Material stiffness Engagement force
1 3 584 45115 182
2 9 665 98470 112
3 10 614 51660 86
4 7 499 27642 170
5 4 453 12617 97
6 6 386 67228 148
7 7 410 89433 64
8 8 539 81441 198
9 3 718 33340 65
10 5 564 59832 127
The figures in the appendix 2 give a clear picture of the bouncing and
performance of the dog clutch for different parameters and different starting
positions. The simulations were carried out with three different models, such as
XX, a dog clutch with patent, the second model by Swepart Transmission AB
and the last model YY, by the thesis team. The tests were carried out for ten
parameter value combinations, see table 1.
Out of these results, simulation number 8 shown in figures 18, 19 and 20, have
much better performance for XX, Swepart Transmission AB and YY. These
results are with the parameter combinations, mass of propeller shaft 8 kg,
rotational speed of the dog clutch 90 rpm, material stiffness 81441 Nm/degree
and engagement force of the male dog clutch 198 N.
23
24
The goal of this thesis work was to analyze three different dog clutch designs using
MSC ADAMS. The first dog clutch was designed by XX, the second by Swepart
Transmission AB and the third by us. We simulated these models in MSC ADAMS
to see which of these models have the best engagement result. The result will be
presented to Swepart Transmission AB.
We consider that this thesis was very interesting since this was a real engineering
concern for Swepart Transmission AB.
We believe in team work and gained a lot of knowledge in MSC ADAMS. It was
interesting to work with classmates from different countries and cultures.
During this tenure we had good communication with our supervisors both at
Linnaeus University and Swepart Transmission AB.
Unfortunately, this ten week thesis period made us to finish off the course early
before seeing the final result, however we look forward to hear from Swepart
Transmission AB about the progress of our result. We believe that we did our best to
achieve our goal and succeeded to get the required result.
The next thesis could be to make the rigid body flexible because, having a flexible
body makes the simulation process more realistic than having a rigid body.
Additionally to continue the designs and simulations by modeling the whole truck as
much as possible in MSC ADAMS.
25
1. http://www.smartmotorist.com/car-accessories-fuel-and-maintenance/four-
wheel-drive.html (2nd of may 2013 at 13:39)
9. http://www.mathworks.se/help/phased/ug/global-and-local-coordinate-
systems.html (2nd of May 2013) kl 16:20
11. Nyagolov, Nelkov, Dimitar, Bashir, Abbas and Genovski, Valentinov, Filip,
(2010) Simulation of the Geometry Influence on Curvic Coupled Engagement,
Master thesis work in Mechanical engineering at Linnaeus University
12. Andersson, Mattias and Kordian, Goetz (2010) FE analysis of a dog clutch
for trucks with all-wheel-drive, Master thesis work in Mechanical engineering at
Linnaeus University
13. http://en.wikipedia.org/wiki/New_product_development#The_eight_stages
(10th of April 2013 at 16:01)
14. http://www.4xforum.com/four-wheel-drive/6-4x4-recovery/using-fwd-
transmission/ (5th of may 2013 at 15:27)
26
http://www.google.se/imgres?imgurl=http%3A%2F%2Fwww.ericsonyachts.org
%2Finfoexchange%2Fattachment.php%3Fattachmentid%3D9143%26d%3D130
9610052&imgrefurl=http%3A%2F%2Fwww.ericsonyachts.org%2Finfoexchang
e%2Fshowthread.php%3F10136-Orienting-a-2-speed-winch-for-Load-
Direction&h=185&w=348&sz=49&tbnid=mX0ym7mCN2YkAM%3A&tbnh=6
8&tbnw=127&zoom=1&usg=__8k_-NC8ZvuonEmn2F-
HVkaVeYLs%3D&docid=Pl3FddG8gbCqJM&sa=X&ei=8zZAUd_QIubh4QTI
_YGABA&ved=0CFEQ9QEwBQ&dur=1829
17.
http://www.core.org.cn/NR/rdonlyres/Aeronautics-and-Astronautics/16-
61Aerospace-DynamicsSpring2003/644258CA-E401-4F96-A505-
E4863A9BBE05/0/lecture7.pdf (27th of may 2013 at 15:22)
27
28
Implementation
After making the required sketch in SolidWorks, the files were converted to
Parasolid (x. t) format.
The first model in MSC ADAMS is shown in figure 1. In this, a blue cylinder
which is marked as (1) is made. One end of the blue cylinder (1) is attached to
the engine and the other end to the transfer gear, marked (2). Transfer gear (2),
rear propeller shaft (3) and front propeller shaft (8) are attached to the ground
using revolute joint. When the dog clutch is not engaged, the power from the
engine will go only to the rear propeller shaft (3) and then to the black rear drive
axle (4). Finally to the rear wheel (5) with the help of cylindrical joint, connected
between rear drive axle (4) and rear wheel (5).The rear wheels (5) can move
along the road because of the contact between the wheel (5) and road (6). The
road (6) is locked with a fixed joint to the ground and the rear drive axle (4) is
connected with translation joint. Now the rear wheel (5) moves because of the
friction. However the middle parts remain stationary.
When the dog clutch (7) is engaged, the front propeller shaft (8) rotates and
finally the front wheel (10). The front transfer gear (11) is to make the same
rotation speed as in the rear wheel (5).
29
30
Figure 1. XX combination 1
31
32
33
34
35
36
37
38
39
40