FULLTEXT02

Master's Degree Thesis
ISRN: BTH-AMT-EX--2016/D06--SE
Design and Structural Analysis of

a Robotic Arm
Gurudu Rishank Reddy

Venkata Krishna Prashanth Eranki
Department of Mechanical Engineering

Blekinge Institute of Technology
Karlskrona, Sweden
2016
Supervisors: R.V Jhansi Rao, Signode India Limited (SIG)

Sharon Kao-Walter, BTH
Design and Structural Analysis
of a Robotic Arm
Department of Mechanical Engineering

Blekinge Institute of Technology
Karlskrona, Sweden
2016
Thesis submitted for completion of Master of Science in Mechanical
Engineering with emphasis on Structural Mechanics at the Department of
Mechanical Engineering, Blekinge Institute of Technology, Karlskrona,
Sweden.
Abstract:
Automation is creating revolution in the present industrial sector, as it
reduces manpower and time of production. Our project mainly deals
around the shearing operation, were the sheet is picked manually and
placed on the belt for shearing which involves risk factor. Our challenge
is designing of pick and place operator to carry the sheet from the stack
and place it in the shearing machine for the feeding. We have gone through
different research papers, articles and had observed the advanced
technologies used in other industries for the similar operation. After
related study we have achieved the design of a 3-jointed robotic arm were
the base is fixed and the remaining joints move in vertical and horizontal
directions. The end effector is also designed such that to lift the sheet we
use suction cups were the sheet is uplifted with a certain pressure. Here
we used Creo-Parametric for design and Autodesk-Inventor 2017 to
simulate the designed model.
Keywords:
Autodesk-Inventor 2017, Creo-Parametric, robotic arm, suction cups.
Acknowledgements
This work is carried out at the Department of Mechanical Engineering,
Blekinge Institute of Technology (BTH), Karlskrona, Sweden and Signode
India Limited (SIG), Rudhraram, India from February 2016 to October 2016
under the supervision of R.V Jhansi Rao.
We wish to express our sincere gratitude to our industrial
supervisors, Signode India limited, India for their competent guidance and
support throughout the project. We are thankful to our academic and internal
supervisor Prof. Sharon Kao-Walter BTH Sweden for her valuable support and
advice.
We would also like to express our deepest gratitude to the staff at
Signode India limited Mr. Madhu, Mr. Annand and Mr. Prashanth for their
timely help, support and everlasting patience. And we would like to thank our
beloved Sir. Sravan Kumar from JNTUH Mechanical department for valuable
discussions and support. We are thankful to SIG in providing us the required
equipment and software to carry out the project.
We are very thankful to our parents for their constant support, love
and care.
Karlskrona, October 2016

2
Contents
1 Notation 7
2 Introduction 9
2.1 Project Statement 10
2.2 Background Study 10
2.3 Research Problem 11
2.4 Scope of Implementation 13
2.5 Objectives 13
2.6 Research Questions 13
2.7 Preliminary Discussion 13
2.7.1 Articulated Arm Robots 14
2.7.2 End Effector of the Robot 15
2.8 Related Works 16
3 Design & Drafting 17
3.1 Mechanical Design 17
3.2 Part 1 19
3.3 Part 2 22
3.4 Part 3 23
3.5 Part 4 & Part 5 24
3.6 Pneumatic Cylinders 26
3.7 Assembly 27
3.7.1 Joint 1 (Waist & Shoulder) 27
3.7.2 Joint 2 30
3.7.3 Joint 3 (Elbow) 31
3.7.4 Joint 4 (Wrist) 32
3.7.5 Joint 5 (End Effector) 33
3.8 Dynamic Behaviour of Robotic Arm 35
4 Materials 42
5 Simulation & Analysis 45
5.1 Need for Stress Analysis 45
5.2 Loads & Boundary Conditions 46
5.3 Export to FEA Module 48
5.4 Meshing 51
5.5 Stress Analysis Environment 52
3
6 Analytical Model 54
7 Results & Discussions 57
7.1 Stress Analysis Results 57
7.2 Convergence 68
7.3 Fatigue Analysis 70
7.4 Stress-Cycle (S-N Diagram) 76
8 Summary &Conclusions 78
8.1 Validation 78
8.2 Conclusion 80
8.3 Future Works 82
References 83
Appendix 84
Link to the Motion Demonstration of the Robotic Arm 84
A. External Forces Acting on Parts 84
B. Deformations on Parts 87
C. Moment Vs Time Graphs 90
D. The Factor of Safety on Parts 92
E. Von-Misses Stress of Part 5 with CFRP material 94
F. Matlab Code for calculating the Number of Cycles 95
4
List of Figures
Figure 2.3.1: The Prototype of robotic arm (Pick and Place Operator). ... 12
Figure 3.2.1: The upper part of Oldham Coupling. ................................... 20
Figure 3.2.2: The shaft and Key of Oldham Coupling. .............................. 21
Figure 3.3.1: The CAD design of Part 2..................................................... 23
Figure 3.4.1: The CAD design of Part 3..................................................... 24
Figure 3.5.1: The CAD design of Part 4 and Part 5. ................................. 25
Figure 3.6.1: The CAD design of pneumatic Cylinder. .............................. 26
Figure 3.7.1: The CAD design of Joint 1 (Waist and Shoulder)................. 28
Figure 3.7.2: The CAD design of Joint 2. ................................................... 31
Figure 3.7.3: The CAD design of Joint 3 (Elbow). ..................................... 32
Figure 3.7.4: The CAD design Joint 4 (Wrist)............................................ 33
Figure 3.7.5: The CAD design of Joint 5 (End Effector). ........................... 34
Figure 3.7.6:The Complete Assembly of Articulated Arm Robot. .............. 35
Figure 3.8.1:The dynamic behaviour of Pneumatic Cylinder. ................... 36
Figure 3.8.2: The Prototype of Robotic Arm in three sections. .................. 37
Figure 3.8.3: The Top view of Robotic Arm. .............................................. 38
Figure 3.8.4: The Graph between Position vs Time. .................................. 39
Figure 3.8.5: The flow chart construction of Robotic Arm......................... 40
Figure 3.8.6: The Graph between Position vs speed. ................................. 41
Figure 3.8.1: The Graph between Strength and Density of the Material. .. 42
Figure 3.8.2: The Graph between Strength and Relative cost per unit volume.
..................................................................................................................... 43
Figure 3.8.3:Part-5 assigned with CFRP material. ................................... 44
5
Figure 5.2.1: The Force acting on the sheet. .............................................. 47
Figure 0.1: The Export to FEA. .................................................................. 48
Figure 0.2: The Part-5 in Export to FEA. .................................................. 50
Figure 0.3: The Output Grapher and Time Series. ..................................... 50
Figure 5.4.1: The Joint were the meshing is excited .................................. 52
Figure 5.4.2: The Meshed Part of FEA. ..................................................... 52
Figure 7.1.1: The graph between Time (sec) vs Stress of Part 1. ............... 59
Figure 7.1.2: The graphs between Time (sec) vs force and moment. ......... 60
Figure 7.1.3: The stress distribution plot of Part 1. ................................... 61
Figure 7.1.4: The Stress distribution of Part 2. .......................................... 62
Figure 7.1.5: The Stress distribution of Part 3 . ......................................... 63
Figure 7.1.8: The FOS of Part 5. ................................................................ 67
Figure 7.2.1: Convergence Plot for Part 2 ……………………………….….69
Figure 2.1.1: Aluminium 6061 Fatigue Data from Experiments ………….72
Figure 7.3.2: The graph between Number of Cycles(N) vs Stress (Pa). ..... 77
Figure 8.1.1: The Factor of Safety for Part 2. ............................................ 79
Figure 8.2.1: The position of Shearing Machine in Industry. .................... 81
6
1 Notation
A Area
B Width
D Diameter of Solid Shaft
E Young’s Modulus
H Height
I Moment of Inertia
L Length
M Mass
Q Weight
R Radial Arm
r Radius of Solid Shaft
t Torque acting on Solid Shaft
T Thickness
V Volume
W Load acting on Solid Shaft
࣋ Density
7
Abbreviations
CAD Computer Aided Design
CFRP Carbon Fibre Reinforced Plastic
DOF Degrees of Freedom
FANUC Fuji Automated Numerical Control
FEA Finite Element Method
FOS Factor of Safety
KSI Kilo pound per Square Inch
SIG Signode India Limited
RCC Remote Censor Control
8
2 Introduction
The most aged methods of metal engaged procedures are
shearing and bending. These are the basic operations that are performed for
metal working. Shearing is a mechanical operation, cutting of large sheets of
metal into smaller pieces of predetermined sizes. When an operation completes
an entire perimeter forming a line with closed geometry is known as blanking.
Shearing machines are of different types, but a typical shear generally consists
of,
• A fixed bed to which one blade is attached.

• A vertically moving crosshead which mounts on the upper blade.
• A series of hold-down pins or feet which holds the material in place
while the cutting occurs.
• A gaging system, either front, back or squaring arm, to produce
specific work piece sizes.
Shearing operation is generally conducted manually, but it can be

conducted using mechanical, pneumatic and hydraulic means also. Currently,
the operation is performed manually at the industry but at a very high risk. The
raw material is collected by the worker and feeding is done into the shearing
machine manually till the sheet is induced completely into it. This operation is
very hazardous to the personnel performing the operation. Also, there is a fair
chance that automating this process might speed up the rate of work when
compared to the manual execution.
To overcome these disadvantages, the entire manual process in
the shearing process is to be automated. In this project, a pick and place
machine is designed to lift the raw material sheets one by one to the shearing
machine. Suction cups are designed as holders for these machines to hold the
metal sheets and place it on the conveyor belt of the shearing machine. This
auto feeding mechanism will be operated by the sheet guide.
This project undergoes an in-depth study of related topics that are explained
in-detail in the future sections. Our main intension is to design this is entire
9
manually operated system (picking of sheet from stack to feed) into
automation, such that it reduces the risk-factor during feeding operation. On
developing this system, we reduce the time of action performed that leads to
increase in productivity.
2.1 Project Statement

The thesis examines the compelling design of a robotic arm i.e. a
pick and place machine and auto feeding mechanism that improves the safety
of the workers. The main intension of designing this pick and place machine
is there will be no need of manual operation of picking the sheet form stack to
shearing machine and the auto feeding mechanism is a continuous process
were the productivity could be effected.
2.2 Background Study

This project is discussed mainly on Design and structural analysis
of a robotic arm, which reduces the man power and might have a good effect
on production rate. This change can motivate the industry and academics such
that the business of the firm is increased. The development in automation can
reduce the revenue cost and raise in capability of delivering the services at low
cost scaling.
To look at the safety of the workmen, we designed a pick and place operator
i.e. a robotic arm and for the feeding mechanism two pneumatic cylinders are
designed. Earlier we have studied about different feeding mechanisms among
those we have designed a new model i.e. using the pneumatic cylinders
pushing the sheet forward through the cutting blades [1]. In this process the
time of feeding is reduced for each sheet.
We have also studied about RCC control for designing the robotic arm[2]. In
this the system integrates manipulator position sensor into the robots control
routine. It also gives the robot its ability to interact with nature. So depending
upon these conditions the manipulator makes it more efficient by providing
self-optimisation system. With this self-awareness of the robot there will be
10
work safety in the environment onsite. Due to this RCC the efficiency of the
manipulator increases. To design these RCC model we need to compare with
revolutionary symmetric structure and circular periodic structure, due to this
we can achieve low stiffness and material will remain same. We have consulted
automation companies like Fanuc Automation Solutions, Rexroth Pneumatic
cylinders and many other information sources searching for most reasonable
and proper solution.
2.3 Research Problem

In sheet shearing operation, picking of sheet and feeding is
undergoing manually which is time taking and risk factor is involved in it. So,
for the first phase (picking of sheet from stack) we need to avoid that by using
an automotive application i.e. pick and place operator and the auto feeding
mechanism is initiated with two cylinders parallel to the sheet. As we have
discussed above the pick and placer is regulated as a continuous operation for
picking the sheets and placing it on the conveyor belt.
Let us consider the pick and place operator i.e. robotic arm design
11
Figure 2.3.1: The Prototype of robotic arm (Pick and Place Operator).
In the above figure 2.3 we need to consider three links of a robotic arm
as the base is fixed. These three links are connected to each other by a finge.
The link1 is having rotary motion, link 2 is having transient movement
upwards and downwards. The link3 is again a rotary motion. Towards the end
of link3 an end effector is being placed which is stable and parallel to the feed.
Through this robotic arm, we need to analyse the forces on individual
component and the complete arm. Now let us consider the free body diagrams
of individual links. The end effector is connected to link3 which is designed
with suction cups to lift the sheet. In this we need to analyse the static and
dynamic properties of the arm. The pressures at vacuum cups is to be converted
to forces and sum of the forces should be greater than the weight of sheet. The
pressure and force relations are to be calibrated for each and every link and
also for the complete arm.
12
2.4 Scope of Implementation
The basic idea in this project is implementation of robotic arm.
Though it can be implemented in various methods, when different parameters
are taken into consideration this model is the most feasible way of
implementation. Other ways of enactmenting the model is enabling them to
adapt to the surroundings.
2.5 Objectives
x Designing, Modelling and Simulation of the pick and place
mechanism. We need to have a time study between currently
undergoing manual operation and newly designed automated
operation.
x The frequency of this operator, it’s repeatability, lifetime etc. are to be
found out.
x The choice of the end effector, it’s design and analysis should be
carried out and documented.
2.6 Research Questions

• How can we automate the process of feeding the metal sheet into the
shearing machine?
• How can we design the pick and place robot to meet the requirements
of the shearing operation?
2.7 Preliminary Discussion

Before designating about the pick and place operators, we have
undergone various methods like use of conveyor belt, pulleys and other
simplified mechanisms for this operation. But after a broad search and enquiry,
we decided to design an articulated robotic system that makes the entire
process more flexible and easier in inducing the mechanism without any
employee payload. The reason for selecting only this particular robot as a
solution depends upon many factors. The very first one is, these sheets get
13
painted and printed one by one and stacked at the end of the printing process.
The transportation of sheets in this process is done using a frame with suction
cups connected to a conveyor belt. The shearing machine is close to the stack
and the direction of the feed in the shearing machine is perpendicular to the
stack direction. It is complex and complicated or in other words not possible
to extend the track of conveyor which is already in use in the printing to deliver
the sheet directly into the feed of shearing machine without stacking. It is
because the time of one cycle of printing and the time of one cycle of shearing
operation of are not same. So, these two process cannot be interlinked without
stacking the sheets. Moreover, the space around these two machines is very
less to adopt any other automation technique. So, keeping these in mind, we
concluded that an articulated robotic arm can do the job of picking and placing
in the given space and can correlate these two operations perfectly.
Our main motive is to reduce the risk factor involved
physically in this operation. So, we took a forward push of designing a three-
jointed robotic system with good and malleable end effector to it. Basically, in
our study the articulated robots are of rotary joint system, that can range from
two to ten jointed and are mechanized by servo motors. They are various
robotic systems, which could be articulated and non-articulated. But according
to our operation we prepared rotary joint system which is articulated, in this
the space consumption is précised as the joints are supported in chain. The
major factors of initiating this system are it is having a continuous path,
acceptable degree of freedom, proper grip, cyclic rotation, good accuracy and
reach, speed control, repeatability and high resolution.
2.7.1 Articulated Arm Robots
Articulated arm robots are generally used to perform risky,

treacherous and highly repetitive and obnoxious works. This entire system is
controlled by a trained operator using a portable device like a teach pendant to
a robot to do its work manually. The main prospective is not the working of
the robot, but how it is to be safeguarded in a regular usage in the industries.
The maintenance depends only on technical operators, how hazardous the use
of the robot system is its environment conditions, position, initialization
requirements, technical errors and other functions.
14
While many engineers working on these robot systems they
could be associated with risks in the operation. In this combination, they need
to use safeguarding methods like repetition and backup systems and the entire
thing should be monitored by a human operator. As the entire system is to be
controlled by an electric device, they are two controllers’ servo and non-servo.
The use of servo controller gives immense feedback about the robot system
and that continually monitors the robot axes which are correlated with the
position, velocity and the entire data is stored in the robot’s memory. Were as
the non-servo controllers do not have the feedback criteria and the system is
controlled through very finite switches. So, in our case as we need the backup
system the servo controller is best to initiate.
2.7.2 End Effector of the Robot
The end effector is one aspect that what brings the robot to
give adaptable solutions[3]. This device is designed to have a great connection
with the environment, and the working of end effector depends completely on
the robot applications. Basically, the end effector is nothing but a gripper or a
device that works according to different applications induced in it and when
we consider it to robotic awareness. They are of different divisions like.
Impactive, Ingressive, Astrictive, Contigutive. These works differently for
different end effectors.
Impactive: - This works as a jaw or fingernail that grasp physically by giving
an explicit collision to the object that is to be acted.
Ingressive: - Use of pins, needles that helps in physically infiltrating the
surface of the object. In my case I use vacuum cups to pick the sheet.
Astrictive: - It is nothing but the suction that is generated on the surface of the
object, which is produced through vacuum cups as outsource and by
electromagnetic stuff if in use.
Contigutive: - It is enforced to have a direct contact for the holding process of
the object, for example surface tension generated at particular point.
15
So, these are the categories based on various physical
belongings of the system. And in individual purpose depending on the working
material like for metal sheets, vacuum cups or electromagnets play a dominant
role as end effectors. In our case, we have taken into the consideration many
factors and took an initiative to design suitable vacuum cups such that the sheet
is picked firmly with a particular pressure exerted and is placed on the
conveyor belt of the shearing operation without any uneven movement of the
sheet.
2.8 Related Works

S. Pachaiyappan, M. Micheal Balraj and T. Sridhar have published a
journal that contains complete data related to articulated arm robots for
industrial applications[4]. They have developed an advanced technique in
working of these particular robots from hazardous conditions and how the
human can intervene into the robotic work zone. The ultimate motive in the
research was to save human lives and in addition of increasing the productivity
and quality of product with good and high technology environment.
Nonetheless they have focused on safety and to create a good and healthy
environment in the industry with use of advanced technology.
After a deep study, we have undergone many methods in development
of articulated arm robots as pick and place operators related to our usage in the
industry. The FANUC system has been taken into reference for developing a
new robot according to the constrained environment and material of work [5].
The entire model is designed in Creo parametric 3.0, the Assembly and
Simulation part is carried out in Autodesk Inventor 2017. As we have
experience and knowledge about these tools earlier.
We have been introduced to robotic arm in JNTU Hyderabad (where
we have earlier done our Bachelor’s), and there we did a detail study on the
manual of robot provided by FANUC. So, we have taken the dimensions based
upon these manuals to develop our model, which are not exactly same
dimensions according to FANUC standards, but have been modified slightly
as we need to do relating to space constraints in SIG (Signode India Limited).
16
3 Design & Drafting
Design of robotic arm means the human supervision on this
operation should be reduced. The shearing operation on which we are working
can handle one sheet at a time. So, the first feature that is expected from this
automation is picking up a single sheet from the stack of many sheets. There
are many options to consider for this carrying operation. For example, a set of
suction cups in a conveyor belt or suction cups replaced with electric magnets.
We cannot use electric magnets as they might pick multiple sheets instead of
one. The cutting blades cannot take more than one blade. Of course, it can cut
those 2 at a time but the blades get worn out quickly. To make sure only one
sheet gets picked up, we use suction cups. If at all we consider the possibility
of making a robotic arm for the purpose of carrying the sheet, then our end
effector should consist of suction cups. Between conveyor and robotic arm,
we chose robotic arm for two reasons. First one was, robotic arm occupies less
space when compared to a conveyor setup. In the industry, we are working, the
place and position of this machine is so important. Just to install an extra
enhancement, the whole shearing machine, which is very huge, cannot be
replaced. So, to get fit into a small space, we thought that robotic arm would
do better than a conveyor setup. The second reason was its portability. How
easily it can be transported. Definitely, a robotic arm can be transported from
one place to another quickly. So, considering all the above reasons, we decided
to design a robotic arm with a pick and place type end effector.
The outline of the total mechanism was drawn on the paper just giving us
the basic idea of the robotic arm which acts as a gateway to our imagination of
how the shapes should be.
3.1 Mechanical Design

The most important aspect and backbone of this thesis is the
mechanical design of the robotic arm. A robotic arm has certain design
specifications and certain parameters are to be taken into the consideration.
Since, the design is an area related to thought, many varieties of designs come
to the mind at the initial stages of the design. Everything might not be fruitful
17
and the trial and error method cannot be trusted blindly. So, keeping all these
things in mind, we have decided to design the robotic arm whose dimensions
are loosely based on the dimension standards of Fanuc robotic arm. The basic
points to be noted and followed for the design are[6]:
Functionality: The arm should have the ability to lift, move, lower and release
an object while closely mimicking the motion of the human arm with full
extension. Any device that can perform the required motions to pick and place
an object required would have met the requirements of this criterion. The
choice of the number of the parts in this particular robotic arm is taken by
comparing it with a human arm. Let the action of human hand picking up a
container appear in your mind. We have the waist, shoulder, elbow, arm, wrist
and fingers do the job. This is the motivation for the choice of the number of
parts. This robotic arm also has 5 parts and 5 joints which are pretty much like
the human hand.
Reliability: The device should be able to consistently pick up and place objects
in a smooth manner. i.e, the motion of the device should be smooth enough to
not drop the objects that are being lifted. Therefore, any device that can lift and
move an object from one place to another without losing any grip would meet
the criteria. After a detailed study, the choice of end effector is made. Since,
this device is used for picking and placing metal sheets, the first common
thought any mind would get is that a magnet can be used to lift the sheet up.
But the problem with that is, the thickness of the sheet is so small that there is
a very high chance of more than one sheet being picked. If more than one sheet
is fed to the shearing machine at a time, that hurts the shearing blade bad which
can reduce the life of the blade. The next option in front of us was to use suction
cups to lift the sheets. This is the most commonly used technique for the
transportation metal sheets in industries all over the world. So, we had decided
to use this technique for this purpose. Also, the industry also had the use of
suction cups and a linear robot (conveyor) for the transportation of sheets in
the printing process. So, we have enough motivation and data to use this
technique.
Motion Range and Speed: Like human body the robots are constructed with
same joints between bones, here we have a constrained limit for the movement
of axis. In our design application, every particular axis has its own capacity of
motion. The degree of movement of robot is calibrated from centre base of
18
axis. By this the speed in pick and place operation might vary, and this is
occurred because each axis moves at different speeds. The complete motion of
the operation is recorded in terms of degrees travelled per second.
Payload: The limited weight of each robot is its payload. So, the critical
specifications and tooling weights are sotted out. In our application, this is
useful in specifying different categories of robots by the above specifications.
Reach: In our articulated robot, we need to check the two extremities that is
nothing but the V-reach and H-reach. Vertical reach is considered to know how
high our robot can go in terms of height extension. Whereas the Horizontal
reach is considered to know the distance of fully extended arm from base to
wrist. In few other applications, we need to even consider a short Horizontal
reach.
Axes: The distinctive segments of our robot are associated with mechanical
joints, that serves as an axis of movement. We have designed our articulated
robot with 5-axis of movement. Generally according to our knowledge
industrial robots are designed to have 6-axis of movement, but the number and
placement of robot just gives flexibility variation for each model.
3.2 Part 1
The first part designed in this project is this. It is because, based on its
measurements we must measure remaining parts. And the weight of all other
parts including the payload will have its great effect on this part since it is the
base of this robotic arm. This part is an assembly of two different parts. Part 1
must rotate on its axis and other parts are connected to this. So, it is the main
source of transportation. The two sub parts in this assembly are upper part and
a shaft. The upper part is designed such that its bottom is the one side of an
Oldham coupling as shown in the figure 3.2 below.
19
Figure 3.2.1: The upper part of Oldham Coupling.
The dimensions of this part are tabulated in the later part of the
document. This part stays in the upper area of the base and will be visible. But
there is a combination of shaft and key attached to this part from the bottom to
which a power source is connected and is made to rotate. This supposed to be
the component that transfers motion (rotational) from the power source to the
upper part of the body. The shaft and key and the total assembly of the part 1
is shown in the figure below. The shaft and key inserted in this assembly acts
as a typical Oldham coupling.
A general Oldham coupling has three flanges, one coupled to the input,
one coupled to the output, and a middle disc that is joined to the first two by
tongue and groove. The tongue and groove on one side is perpendicular to the
tongue and groove on the other. The middle flange rotates around its centre at
the same speed as the input and output shafts. Its centre traces a circular orbit,
twice per rotation, around the midpoint between input and output shafts. For
this operation, we modified the Oldham coupling a bit. Instead of using 3
flanges, we used only two, which are not flanges exactly. We designed the ends
of the 2 parts in the base as Oldham couplings as shown in below figure 3.2.2.
Since this is a pick and place operation, two disc instead of three might
be able to withstand the torque ranges of this operation. As we can see the
lower part of Oldham coupling has a shaft and a key. The upper end of the
20
coupling has a flange which is directly designed with shoulder of the robotic
arm. All the assembly described till now will be assigned a rotatory motion.
That means, a motor is attached to the shaft at the lower most part in the figure
below and if it rotates, the whole robotic arm rotates.
Figure 3.2.2: The shaft and Key of Oldham Coupling.
The dimensions of the shaft are given below.

Length of the shaft (L) = 0.182 meters
Radius of the shaft (r)= 0.12 meters
Volume of the shaft (V) = 0.002055 meters3
21
3.3 Part 2
This part can be compared to the bicep of the human arm. Like
the muscle of the human arm, it moves very less when compared to the forearm
but it provides the strength and hold for the forearm and wrist to do their jobs.
Its length is supposed to be more than part 1 and less than part 3 (since our
design is like the human arm). An end of this arm is connected to part 1 creating
joint 2 and the other end is connected to the part 3 to create joint 3. The design
of this arm is shown below. The dimensions are tabulated in the later part of
document. The upper end of the part is having a shaft to make itself rotate
around the axis of the joint between part 1 and part 2. The lower end of this
part is having holes and a gap between the two extended grooves to place the
part 3 creating joint 3. The power source is connected to the shaft at the top
end part 2. The rotation of this arm happens at this end. Its assembly, creating
the joint and its limits are discussed in the assembly section. The dimensions
of this part are given below.
Length of the part (L)= 0.45 meters

Thickness of the part (t) = 0.08 meters
Width of the part (b)= 0.153 meters
Volume of the part (V)= 0.00359787 meters3
Area of the part (A)= 0.269904 meters2
22
Figure 3.3.1: The CAD design of Part 2.
3.4 Part 3
This part is the forearm of this robotic arm. The reach of this
robot mainly depends upon this part. This is a bit longer than part 2. The
movement of this part is more when compared to the other parts. This part with
part 2 creates a joint which is like the elbow in the human arm. The main
purpose of these two arms is sustaining the weight that is lifted by the arm.
One end of this arm is connected to part 2 as describe earlier. The other end is
connected to part 4 which is a kind of wrist to this hand. The power source for
this arm is given to the shaft on the left end in the picture below. The
dimensions of this part are:

Thickness of the part (t) = 0.08 meters
23
Figure 3.4.1: The CAD design of Part 3.
3.5 Part 4 & Part 5

Now we have arrived at the wrist part of the hand. Part 4 is the
linkage itself between end effector and part 3. It has the ability to rotate around
the axis at the end of part 3 (around X axis). It can rotate 360 degrees on the
axis. How much it should rotate can be adjusted as per the requirement of the
user. It also holds the end effecter (Part 5) in the desired position. It acts a joint
for the part 5 where part 5 can rotate around its own axis (Y axis). The part 4
is shown in the figure 3.5 below on the left. Joint 4 is created between these 2
components. Part 4 being the stable one, it lets part 5 rotate around Y axis.
Length of Part 4 (L)= 0.266 meters
24
Height of Part 4 (H)= 0.155 meters
Volume of Part 4 (V)= 0.000875211 meters3
Area of Part 4 (A)= 0.133435 meters2
On the right of this picture, the part 5 can be seen. This is the one that rotates
around Y axis. This part allows pneumatic cylinders to get connected to it. This
part has the capacity to sustain the weight of 8 pneumatic cylinders and the
sheet attached to them that is to be lifted. This part has the space to carry 8
more pneumatic cylinders. The dimensions of this part are:
Height of the part (H)= 0.515 meters
Figure 3.5.1: The CAD design of Part 4 and Part 5.
25
3.6 Pneumatic Cylinders
As we discussed earlier, the means of holding the sheet in this
operation is executed with the help of suction cups and pneumatic cylinders.
The phenomenon of this action is that; rubber suction cups get places on the
sheet. And when ready, the air between the suction cup and the sheet is sucked
out and a pressure near to the vacuum is created. The pressure outside the
suction cup in way too larger than the pressure inside the cup. So, the air tries
to enter inside the suction cup through the gap between the sheet and suction
cup. This automatically creates an air lock and the sheet gets strongly attached
to the suction cups. This phenomenon is already in use in the industry in the
printing process which is already mentioned earlier. The pneumatic cylinder
assembly is shown below.
Figure 3.6.1: The CAD design of pneumatic Cylinder.
26
3.7 Assembly
After the design of individual part is developed in CREO Parametric
3.0 the next step is we need to assemble the parts to form a complete robotic
arm using Autodesk Inventor 2017 Software[7]. The assembly has the arm
with the wrist and end effector. If we describe the functionality of the robotic
arm, it is a six degrees of freedom system. Six degrees of freedom (6DOF)
refers to the freedom of movement of a rigid body in three-dimensional space.
Specifically, the body is free to change position as forward/backward (surge),
up/down (heave), left/right (sway) translation in three perpendicular axes,
combined with changes in orientation through rotation about three
perpendicular axes, often termed pitch, yaw, and roll. The part in the yellow
rotates around its axis. The parts in red and brown are fixed at their bottom
ends and move up and down. Now, the most important area of the robotic arm
is described. It is the wrist and end effector.
And the dynamic behaviour of these assembled joints is explained
below.
3.7.1 Joint 1 (Waist & Shoulder)
In the below figure 3.7.1, we can see 3 parts. The base is

fixed which is obvious and the first part was coloured in black. Part 1, as
marked in the figure 3.7.1, can rotate on its axis perpendicular to the base. How
much it to rotate that must be decided by the user. The base does not rotate by
itself. We use a power source which rotates the base. The base we designed
might not be as simple as the outline shown below. The base area we designed
consists of 3 parts. First one is the fixed, which provides a fixed support to the
remaining moving arms. The second and third parts are the rotating ones. They
are together assembled as shown on the figure 3.7.1. So, first the base part is
to be fixed firmly to the ground and joint 1 is between the base and part1, the
nature is it rotates around its common axis at a fixed limit of 0 to 360 degrees.
The distance between part 1 and base is 1/8th of an inch and it is generated to
avoid the friction between them, but we can set the gap between these two parts
by an option in inventor. In the figure 3.7.1 below, the left one has a yellow
arrow mark showing the direction of rotation and
27
Figure 3.7.1: The CAD design of Joint 1 (Waist and Shoulder).
As we know the dimensions of the solid shaft the length, volume and radius,
now to find the torque of the solid shaft we need to consider few other
dimensions.
R= 0.958 meters
W= 33N
Where,
R is the distance of the radial arm from centre of shaft to maximum bending
length of the arm.
W is the load acting on the shaft.
So, by considering the above dimensions the torque is to be calibrated.
28
Torque T= load acting on the shaft is multiplied to distance of the radial arm
from centre of shaft to maximum bending length of the arm.
Therefore, the torque is T= W * R

T= 33 * 0.958
Torque of the solid shaft is = 31.614 Nm.
To know the stability of the shaft we need to find the FOS that is nothing but
the Factor of Safety of a solid shaft.
We need to calculate Induced shear and allowable shear to find the Factor
of Safety.
்௢௥௤௨௘‫כ‬ଵ଺
Induced Shear = గ‫כ‬ௗయ
Here, as we know the radius of the shaft is r = 0.12meters

Then the diameter is d = 0.24meters
Now, we need to substitute the calculated values in the formula above to find
the induced shear of the shaft.
Therefore, Induced shear is = 31.614 * 16 / 3.14 *(0.24)3
= 505.824 / 0.0434
= 11654.9 KPa
Induced shear is = 11.65 MPa
As we know the induced shear the allowable shear is to be taken form the
ASME code depending upon the material we have considered.
So, according to ASME code for T6 6061 Aluminium material the allowable
shear of solid shaft is calculated below.
The maximum shear stress should be 0.3 times tensile stress as per ASME
code.
29
For T6 6061 aluminium alloy the tensile = 276 MPa
For T6 6061 aluminium alloy the UTS = 310 MPa
We must take the highest value that is UTS for calculating the allowable
shear as per ASME code.
Allowable shear = 0.18 * 310
= 55.8 MPa
Therefore, the factor of safety is = allowable shear / Induced shear
= 55.8 / 11.65
The Factor of safety of a solid shaft is = 4.75
3.7.2 Joint 2
This is the joint that is formed by the combination of Part 1

and Part 2. Part 1 is associated to base with a joint (joint 1). This particular
joint allows part 1 to rotate on its axis (Y axis). Now as part 2 is attached to
part 1, part 2 can rotate on its axis (Z axis) in an up and down movement, and
can rotate around the axis of part 1 (Y axis) simultaneously as both of these
parts are connected. The anatomy of this joint is shown in the figure 3.7.2
below.
30
Figure 3.7.2: The CAD design of Joint 2.
The joint 2 has a fixed limit of rotation where part 2 (the part in
yellow) rotates from 55 to 270 degrees. However, these limits are given such
that both the parts consisting this joint do not collide. These can always be
changed by the user. The gaps between walls of part 1 and part 2 are equal,
and this is developed by avoiding all the possible contact among them, such
that there is no proper friction development and the gap is also adjustable.
3.7.3 Joint 3 (Elbow)
The further anatomy of the robotic arm deals with the elbow. The
elbow is the joint of two parts 2 and 3. The main purpose of this elbow is to
give the arm some more room to move the end effector forward. Although each
link in a robotic arm is important and must bare some weight, but this elbow
is quite important as it can be termed as the centre of the robotic arm. The
elbow is a joint of two parts which are shown below.
31
Figure 3.7.3: The CAD design of Joint 3 (Elbow).
These two parts are assembled to each other and their common point
becomes the joint. Both of these are supposed to rotate around Z – axis i.e.,
they should move up and down. The yellow part shown in the figure is
connected to the shoulder of the robotic arm shown above. This yellow part is
continued with another arm in red as shown in the figure 3.7.3. Joint 3 has a
fixed limit of rotation that is the red part moves on positive axis from 0 to 100
degrees and negative axis from 0 to –120 degrees.
3.7.4 Joint 4 (Wrist)
Wrist is the area which balances the operation. In this operation,

the most important point is to balance the sheet while it is being transported.
This robotic arm must pick a sheet and place it in another point. The picture
below shows the top view of a prototype of the robotic arm. In this picture,
there is a starting point and end point. The sheet is carried all the way from
pick up point to end point. The most important point is, the sheet must remain
parallel to the shearing surface all the time. The wrist plays an important role
in doing this. The construction of this joint done in such a way that part 4
rotates around its axis (Z axis) and hold the end effector. Since part 4 is smaller
in size when compared to part 3, it is important to balance the part 4 between
32
the groves of part 3. Improper balance might affect in non-uniform loading on
part 3 and uneven deformation.
Figure 3.7.4: The CAD design Joint 4 (Wrist).
Joint 4 has a fixed limit of rotation that is positive axis from 0 to 230
degrees and negative axis from 0 to –20 degrees. In the part 2, 3 and part 4 are
having joints in between them which are identical and its nature doesn't
change.
3.7.5 Joint 5 (End Effector)
Joint 5 is the most important and delicate part of this whole

assembly. This joint is between part 4 and part 5. Part 5 rotates around Y axis
at the joint. Part 5 again needed to be assembled with the pneumatic cylinders.
The movement of joint 5 is identical too joint 1 and hence the limits are also
the same that is 0 to 360 degrees. In this the joint 1 and 5 are moving around
y-axis and joints 2, 3 and 4 are moving around z-axis. In joints 6 we have eight
pneumatic cylinders on the end effector as shown in figure 3.7.5, which means
they consists of eight pistons and each piston is set into its respective cylinders
with the same limits. The limits are (if the piston wall is in touch with topmost
33
wall of pneumatic cylinder and the movement is possible only downwards),
the piston has a limit of 2.5 inch of maximum reach as 0.5 inch as the end, the
same thing happens for the remaining seven pistons and thus the articulated
arm is assembled. The figure 3.7.5 below is the full end effector which has 8
pneumatic cylinders and suction cups at its ends.
Figure 3.7.5: The CAD design of Joint 5 (End Effector).
All these joints assembled together gives the robotic arm, a pick
and place operator. This robotic arm is now consisting of parts those are
movable. Now, we must assign jobs for these parts. The main job is to pick
and place a sheet metal from one position to another. This main job is divided
into smaller jobs and assigned to each part and joint. Them working
simultaneously as per directions given gets the job done. These set of directions
explain dynamic behaviour of the robotic arm.
The complete Assembly of the Articulated Arm Robot is show in Figure 3.7.6.
below.
34
Figure 3.7.6:The Complete Assembly of Articulated Arm Robot.
3.8 Dynamic Behaviour of Robotic Arm

The very first job that is needed to be done is holding the sheet and
getting ready to the take off. This work is done by the end effector. In our case,
pistons of 8 pneumatic cylinders we have, shoot down and suction cups get
themselves attached to the sheet. When they get attached to the sheet and an
air lock is created between suction cups and the sheet, the piston rods retrace
themselves into the cylinders creating a gap between ground or stack and the
lifted sheet. This retracing action takes place in half a second, were this time
is not a fixed. The user can edit the time as per his wish and requirements.
These inputs are given in the dynamic simulation module[8] in Autodesk
Inventor 2017. All proper assemblies created in Inventor are converted into
required joints in the dynamic simulation module. For example, the piston is
placed in the cylinder and constrain its motion limits such that it moves in
between front and rear wall of the piston. This assembly is converted into a
cylindrical joint in the dynamic simulation environment. Cylindrical joint has
a nature where the part attached can move translationally and can rotate on its
axis. This rotation is unwanted for our use here. Only translational motion is
35
desired. Autodesk Inventor provides an option to lock the DOF’s using which
we lock the rotational motion in this part. The total action of lifting the sheet
up starts at 0 seconds and ends at 0.5 seconds. Now the next motion starts from
0.6th second.
Figure 3.8.1:The dynamic behaviour of Pneumatic Cylinder.
The next job is to lift the sheet up to a safer height, so that it can be
transported from one place to another. The word safe is used because, if it is
not at a proper height the sheet might hit and collide with other machinery or
may be any person standing near. This job of lifting it up is a result of
simultaneous actions between part 2, part 3 and part 4. At 0.6th second, part 2
starts moving upwards (Anti-clockwise direction). By the end of 2.5th second,
part two changes its position by 37 degrees. During this time period, there is a
simultaneous movement in part 3. It also moves upwards changing its position
by 10 degrees and then 20 degrees downwards. The reason of this up and down
movement is that, there is a safe distance between end effector and the other
parts of robotic arm. Collisions are messy but the movement of part 3 starts
36
from 0.7th second. All these time limits and commands can be changed as per
the user.
Now the most important task arrives, that is to try maximum how to
maintain the sheet parallel to the ground (machining surface). The reason is,
since the sheet is held based on an air lock between suction cups and the sheet.
So, if the sheet is slant, it is forced to pull itself downwards and air lock can be
broken. So, to avoid this and to maintain the hold, it is important to maintain
the sheet parallel to the ground.
Figure 3.8.2: The Prototype of Robotic Arm in three sections.
Let us consider a human wrist while lifting and carrying a tray with
glasses full of water in it. While lifting the tray, even though the arm moves
upwards the position of the wrist changes with the motion to keep the tray
parallel to the ground, so that the water won't spill. The same mechanism is
being used here in the wrist. If we see the picture above, it is a rough sketch
demonstration of the current action. We observe that the blue part and the
sheet are parallel to the machining surface all the time. That blue part is part 4,
which is described in the earlier sections. As part 2 and part 3 moves upwards
or downwards, part 4 rotates in the opposite direction to put this sheet parallel
to the ground. For this action from 0.6th second to 2.5th second, the part 4 to
which end effector is connected rotates 28 degrees in clockwise direction.
These three simultaneous actions lift the sheet up to the desired height and
maintaining it parallel to the ground.
37
The next motion that follows is to transfer the sheet from its position
to the destination. This is a solo action of joint 1, the rotational joint between
base and part 1. The part 1 rotates and all other parts connected to it displace.
This rotation starts at 2.5th second and ends at 5th second. The part 1 changes
its position by rotating 90 degrees clockwise from its current position. But that
is not the only task here. Let’s observe the rough sketch below. It shows three
steps of this action. In all these steps, the circle in the figure is the top view of
the part 1. In section 1, the robotic arm is at its 2.5th second i.e. the starting
point of rotation of part 1. Section 2 is its position during the travel and Section
3 shows the end point of this rotation (5th second). In all these three pictures,
the sheet which is light green in the picture, always stayed in the same position
with respect to the red line in the picture during the transportation. It only
changes its position with respect to ground. This action is executed by the joint
5 (Rotational joint between part 4 and effector). As joint 1 starts acting, joint
5 also starts acting. The end effector rotates from 2.5th second to 5th second by
90 degrees, but in counter clockwise direction.
Figure 3.8.3: The Top view of Robotic Arm.
This particular action is assigned in this process only to test and verify
the abilities of the robotic arm. These can always be changed by the user. As
the sheet reaches the 5th second, the next task is to put it down. This action is
again related to part 2, part 3 and part 4. This action is a mirror image of the
38
action of lifting up performed by these parts (0.6th second to 2.5th second) only
difference will be the timing of this action. This action starts from 5th second
and ends at 7th second. Again, the suction cups shoot down and put the sheet
down. This action spans from 7th second to 7.5th second. As an example, the
position vs time graph of part 2 is shown below. The whole action spans are
for 7.5 seconds. The steps involved in assigning these positions and speeds at
each joint is explained next.
Figure 3.8.4: The Graph between Position vs Time.
In environments section in Autodesk Inventor 2017, there is

the dynamic simulation module. Once it is opened, we can see a section called
standard joints in model tree. This is obtained while performing the assembly.
All joints that are assigned in the assembly are taken as standard joints. In our
case, all the joints described above are present here. If these are unwanted or
needed to be modified, there is an option available in this module. In bar the
upper part of the screen, we find some option simulation settings. On clicking
this, it gives an option Automatically convert constrains to standard joints. By
clicking that off, we can edit our own joints in dynamic atmosphere using the
39
insert joint option in the upper part of the screen. Now it’s just assigning the
actions for every arm and coordinating their actions according to the job that
is needed to be done. For example, select a joint in our case its joint 3 (Part 2
and part 3). Find this joint in standard joints section in model tree. Right click
on it and select properties. Find edit DOF’s option and start assigning the
values either a numerical value or an input grapher. The process that is done
for this joint is shown in a flow chart 3.8.5 below. Following these steps lead
you to the command box shown in the following figure. We have the choices
to edit the DOF’s, change the path of a part or lock the DOF’s such that the
parts do not move at all. In this assembly for robotic arm, we have used
rotational joints for joint 1, 2, 3 and 4 and we have used cylindrical joints
between piston rods and pneumatic cylinders.
Now, let us see the construction of robotic arm in a flow chart below.
Dynamic
Environm Standard
Simulatio
ent Joints
n
Edit Revolutio
Imposed Properties n (Part2
Motion &Part3)
Input
Position
Grapher
Figure 3.8.5: The flow chart construction of Robotic Arm.
40
Figure 3.8.6: The Graph between Position vs speed.
41
4 Materials
Robots are mostly built of common materials. Some specialized
robots for clean room applications, the space program, or other "high tech"
projects they may use titanium metal and structural composites of carbon
fibers. The operating environment and strength required are major factors in
material selection[9]. There are a wide variety of metals and composites
available in the market these days. Selection of material is very deep process.
We have referred material and process charts designed by Mike Ashby[10]. He
has provided us with a wide range of plots showing the different qualities and
characteristics of materials plotted against each other. Of all them, we focused
on 2 chats, strength vs density plot and strength vs relative cost figures 4.1 and
4.2. Selection of materials and the cost study to design an economic model is
a completely different and deeper area of engineering. That case study requires
more parameters to compare judge the choice of materials. We are not getting
into that now but, we tried to choose the materials in such way that they satisfy
our load bearing capacity requirements and not too expensive.
Figure 3.8.1: The Graph between Strength and Density of the Material[10].
42
Figure 3.8.2: The Graph between Strength and Relative cost per unit
volume[10].
Steel, cast iron and aluminium are most often used materials for the
arms and bases of robots. Aluminium is a softer material and therefore easy to
work with. But steel is several times stronger. We chose to design the parts
with aluminium. It hasn’t been a direct selection. The part 1 has a choice to
have more weight when compared to other parts. The remaining parts are not
grounded and needed to carry some weight of other arms and end effector also
the sheet in this case. So, they cannot be heavier than the base which can harm
the part 1. We decided to assign aluminium to all parts including part 1. Of
course, we always have choice to change the material. Our selection has been
proven good as aluminium did good for parts 1,2,3 and 4. But, in the opposite
case, we decided to assign stainless steel to the parts. Only part 5 showed less
load bearing capacity with aluminium but, we have tried changing the material
to CFRP (Carbon Fiber Reinforced Polymer). The reason for this is it is
stronger and lighter material than aluminium. We cannot assign a material with
more density. As density increases, weight increases and increase in weight of
Part 5 might have bad effects on other arms. Our experiment worked and the
part 5 has increased its capacity in figure 4.3. The only problem with CFRP is
it is relatively expensive. But again, selection of materials is a completely
different study. We limit our study to the usage requirements of the industry.
43
The material used for most of the parts of robotic arm is Aluminium 6061
alloy[11]. Aluminium 6061 is a precipitation-hardened aluminium alloy,
containing magnesium and silicon as its major alloying elements. It has good
mechanical properties, exhibits good weldability, and is very commonly
extruded (second in popularity only to 6063). It is one of the most common
alloys of aluminium for general-purpose use. The mechanical properties of
6061 depend greatly on the temperature, or heat treatment, of the material.
Young's Modulus is 69 GPa (10,000 ksi). Annealed 6061 (6061-O temper) has
maximum tensile strength no more than 120 MPa (18,000 psi), and maximum
yield strength not more than 55 MPa (8,000 psi). The material has elongation
(stretch before ultimate failure) of 25–30%.
Figure 3.8.3:Part-5 assigned with CFRP material.
44
5 Simulation & Analysis
5.1 Need for Stress Analysis
A good design is always judged by its load bearing capacity. In
operations like transport where machines are used, we expect them to be
stronger than the necessity, so that they can perform the required action for a
long time. Stress analysis is an engineering discipline that uses many methods
to determine the stresses and strains in materials and structures subjected to
forces. In engineering, stress analysis is often a tool rather than a goal; the
ultimate goal being the design of structures and artifacts that can withstand a
specified load, using the minimum amount of material or that satisfies some
other optimality criterion. The stress on the part can be related with the load
carrying capacity. Typically, the starting point for stress analysis is the
geometrical description of the structure, the properties of the materials used for
its parts, how the parts are joined, and the maximum or typical forces that are
expected to be applied to the structure. The result of the study (analysis) is a
description of how the applied forces spread throughout the structure, resulting
in stresses, strains and the deflections of the entire structure and each
component of that structure. The analysis may consider forces that vary with
time, such as engine vibrations or the load of moving vehicles. In that case, the
stresses and deformations will also be functions of time and space. Generally,
the calculation of the stress distribution used to be carried out using
mathematical means and Finite Element Methods. But calculation of stress
distribution more appropriately needs more number of elements in the
geometric design. This increases the computational time and effort by doing it
only in the mathematical way. Hence, we take the aid of simulation techniques
of CAD software’s such as Autodesk Inventor 2017 to perform these tasks and
save the effort. To establish a good simulation environment, we must apply
proper loads and boundary conditions to make the model as close to reality as
possible.
45
5.2 Loads & Boundary Conditions
The robotic arm has 5 joints and every joint has only one degree of
freedom (rotational). To do this, we must specify an axis for each joint to rotate
around it. Any other kinds of movement are constrained. The way the joints
are created are explained in the Assembly section in the earlier part of the
document. In this area, we discuss about the Behavior of the joints under loads
and boundary conditions. The figure 3.7.2 shows joint 2 of the assembly. The
joint has already been defined as the rotational joint. Now, for part 2 to rotate,
a moment must act on the shaft of part 2. The angular momentum of an object
can be connected to the angular velocity ω of the object (how fast it rotates
about an axis) via the moment of inertia I (which depends on the shape and
distribution of mass about the axis of rotation). The part 2 moves from 230
degrees to 193 degrees in anticlockwise direction for 37 degrees in two seconds
giving the angular speed of 18.5 degrees/sec. The speeds of the joints are
tabulated below table 5.2 for the calculation of moment.
Table 5.2.1: The table for Velocities of all Parts.
Part No. Start point End point Time taken Velocity

1 360 degrees 270 degrees 2.5 secs 36 deg/sec
2 230 degrees 193 degrees 2 secs 18.5 deg/sec
3 168 degrees 178 degrees 5 secs 20 deg/sec
4 186 degrees 214 degrees 1.2 secs 23.3 deg/sec
5 360 degrees 270 degrees 2.5 secs 36 deg/sec
The angular moment L can be calculated using the formulaࡸ ൌ ࡵൈ࣓,

where ࡵ is the moment of Inertia of the body and ࣓ is the angular speed.
46
Inventor performs these calculations itself and takes the moment value for the
calculation. Along with the speeds of the joints, we also give the loads on the
end effector. The sheet is an Iron sheet with measurements 777ൈ400ൈ4 mm,
mass of the sheet is 3.35 Kg and weight 32.9286 N. There are 8 suction cups
on the end effector. The weight of the sheet acts downwards (due to gravity).
The suction cups need to create a force that overcomes gravity. The only
external loading given in this model is, the pressure acting on the sheet at the
suction cups to lift the sheet up. As we said earlier, the suction cup
phenomenon is already in practice in the industry. There the pressure created
on the sheet in the direction against gravity by the locking of suction cups is 3
bar. In our case, we have 8 suction cups. This pressure is divided among 8 cups
and is converted to force by multiplying the divided pressure with the area of
the suction cup.
The overall pressure on the sheet is, P = 3 bar = 300000 Pa
Divided among 8 cups, pressure by each cup is, Pcup = 37,500 Pa
The area of the suction cup (Calculated from Creo) is, Asheet = 0.00108 meter2
Now, the force created at each cup is, F= Pressure x Area = 40.5078 N
The force F that is calculated above is the force that acts against gravity at one
suction cup. There are 8 of them on the end effector so the net force against
the gravity is 8 times F.
Figure 5.2.1: The Force acting on the sheet.
47
The remaining loading in this model is automatically taken from the assembly.
The external loads are to be given separately.
5.3 Export to FEA Module

The dynamic simulation module only provides the means to learn and
correct the dynamic behaviour of the assembly. We have provided the loads,
forces and constraints needed for the robotic arm. There is a particular timing
for each part and all of them moves correspondingly with each other. All the
action happens in 7.5 seconds and the movement is solved for stress in stress
analysis module.
As discussed earlier, all these components are to be exported to FEA to
perform stress analysis. For this we have export to FEA option in Dynamic
Simulation module. All that is needed to be done is selecting the part that is
needed to be exported. Click on the export to FEA, then select a part or multiple
parts. As soon as the parts get selected, a window opens asking to select all the
joints and contacts related to it. For example, part 2 is selected to export. The
links we have for part 2 are joint 2 and joint 3. We select the surfaces of part 2
that has direct contact with joint 2 and joint 3. Results can go really bad. The
export to FEA module and selection for part 2 are shown in the figure below
0.1.
Figure 5.3.1: The Export to FEA.
48
Part 2 doesn’t have any welded joints attached to it. Even though
the selection is same for those components with the welded joints, here is the
example of part 5 which is associated with joint 5 and has 8 pneumatic cylinder
surfaces welded to it. The selection is shown in the figure below. It is important
that the component that is to be exported must be a flexible one. That means,
sometimes for our convenience, we assemble different components in different
assembly atmospheres and join them under one atmosphere later. The final
assembly which has sub-assemblies in it treats all the sub-assemblies as parts.
So, to select a single part from a sub-assembly, we need to select that sub-
assembly and right click on it, then select the option flexible on it. In this way,
we select all the necessary parts that are exported to Stress analysis module in
Autodesk Inventor.
After selection, run the simulation using the simulation player. In this
case, the total time period is 7.5 seconds. After the simulation, has run, open
the output grapher in the upper part of the screen. There we find the export to
FEA option. Click that and we can see the parts that are selected for FEA. Now,
we must generate a time series. Right click on the time series option and select
generate series option. A window opens up showing the number of steps
present between 0 and 7.5. We can select any number of time steps according
to which we are going to analyse those parts. In the output grapher itself, we
have all the stable groups, welded groups and standard joints present in it. Click
any of the joint, select any option present there such as position, velocity and
acceleration of that joint. Run the simulation and you can see the path of that
particular physical quantity throughout the run time of the simulation. Which
brings us to the point, the velocities or torques that are assigned for the joints
with which they move.
49
Figure 5.3.2: The Part-5 in Export to FEA.
Figure 5.3.3: The Output Grapher and Time Series

Now, exit the dynamic simulation module and go to environment
module. Go to stress analysis. Create the study, select the option motion loads.
This option gives us the chance to select a particular part at a particular time.
Here, for example, we chose part 2 to be simulated at the time 1.5 th second.
Select that time or any time step closer to this value. Click ok and we enter the
stress analysis environment. Here, the part which is going to be simulated
appears as it is but all other parts in the assembly fades out. We can also see
the yellow arrow marks all over the figure. Each arrow mark represents a load
or force acting on that body. All the data is taken from the information
50
generated by the dynamic simulation. Now, we assign the materials for the
parts here. Click materials section, go to assign materials and give each part a
material which is desired. Here, we give aluminium for all parts except suction
cups and sheet metal. Suction cups are given the rubber material and sheet
metal is assigned with cast iron. The density of the material multiples itself
with the volume of the part automatically giving the mass of the part. As the
gravity is also assigned, the weight of the component is also added to the loads
section. All this work is done by the software itself. The loads are given;
constraints are set and materials are assigned. The next step is meshing. Go to
the mesh controls, edit mesh size and click mesh view. The mesh is generated.
Now, check for geometry updates such as contacts. Update contacts if
necessary and run the simulation. Now, we arrive at the most interesting part,
interpretation of results.
5.4 Meshing
The meshing of the designed articulated arm robot is accomplished
very cautiously to achieve good amount of stress distributions. The stresses
generated here are very high and these stresses are developed due to bending
loads at the round and sharp corners. And by having a deep study about
articulated arm robot cautiously, the high stresses generated are calibrated and
are characterized as detracting ranges.
By default, meshing provided by inventor is made and the part is
simulated, after the detail study of this stress distribution obtained. We found
some areas were stresses are more when compared to remaining areas. The
initial sizes of mesh are 0.1 minimum element size and the average element
size is 0.2. Then the local mesh area option present in inventor to select the
areas those are observed to possess high stresses and their element dimensions
are changed to 0.01 minimum element size and 0.02 average element size. The
local mesh control gives us the option to edit the density if mesh at any chosen
area. The mesh adjusts itself according to the shape of the geometry. The shape
of the element is tetrahedral and during the final stages of simulation, we made
sure that each geometry has at least 2 lakh elements. The meshed Finite
element model is shown in Figures below 5.4 and 5.4.1.
51
Figure 5.4.1: The Joint were the meshing is excited
Figure 5.4.2: The Meshed Part of FEA.
5.5 Stress Analysis Environment

The Finite element analysis in Autodesk Inventor becomes really easy
when the assembly is proper and the joint forces, loads and constraints are
52
given properly. The joints and parts are described and the way they are
assembled is explained in the previous section. In this section, we discuss about
the finite element analysis performed by the Autodesk Inventor on our robot.
For this, we have exported the parts from dynamic simulation module to stress
analysis module[12]. The movement of each and every part of robotic arm is
preloaded and exported to FEA. The constraint set is very important for any
structural analysis of any physical structure. If we have a load acting on a body,
there must be a reaction of that load on that body. Physical constraints
determine that reaction. For example, let us consider a cantilever beam with a
load acting on its free end. Naturally, the deformation will be more at the free
end. But its reaction is learnt at the constrained end. So, the strength and
stiffness of that beam for that material is learnt.
Similarly, we constrain the components of the robotic arm
according to our use. Each arm in the robot is involved in at least one joint.
Even though the connections won’t change during the process, the orientation
of stresses on the components change due to their motion in the given time.
Inventor provides the option of motion loads in stress analysis module. Where
we analyse each part at any interval of time including the external loads, body
loads and gravity. We just must generate a time series with the number of time
steps between them. In this case, we have taken 25 time steps between 0 and
7.5 seconds. We choose 6 time points where we feel the change in the stresses
and displacements can be observed clearly. The loading in this model is
automatically taken from the assembly. The stress is analysed dynamically as
we must know how the stresses are fluctuating in a part during its motion. The
loads and boundary conditions change and recorrect themselves whenever
there is the change in the position or time step. Now, we will have a stress
distribution once the simulation is done. But, how do we know whether the
results obtained are correct or at least approximately correct? Even if they are
correct, are the results good enough to say that the design can take the load.
Let’s have this discussion in the coming sections.
53
6 Analytical Model
The hand calculations are performed in this case are just to verify
whether the simulation is going according to solid mechanics theory. Since it
is very tough and complicated to establish a 3D model with hand calculations,
we always prefer simulation software’s of FEA (Like Autodesk Inventor
2017). But to verify whether the given boundary conditions and created
environment is correct, we tried to establish a deformation equation for a part
in our model. Here we are going to compare the calculated value to the
displacement (deformation) in x-direction in the simulation.
For this purpose, we take part-2 to carry on the hand calculations
with in. we have few changes in the geometry of this part, such as we have
ignored gaps and holes in the geometry and haven taken it has a solid linear
beam with uniform cross section. The curved cross section is taken has a
straight line. The dimensions of part-2 are given below.
Volume V = 0.00359787 meters3

Density ߩ = 2700 kg/meters3
Mass M = volume * density
= 9.714 kg
Weight q = mass * g = 95.29N
Length L = 0.45 meters
b = 0.153 and h = 0.08
Euler-Bernoulli Beam theory

Euler-Bernoulli beam theory is a simplification of the linear theory of
elasticity which can calculate load carrying and deflection characteristics of
beam. We start with a 4th order differential equation of deflection and derive
the 1st order equation. For this purpose, we take the position of beam at 0.5th
second and assume it as a simply supported beam, with both ends fixed[13].
The boundary conditions of a simply supported beam are.
54
x w (0) =0. Because the beam is pinned to its support, the beam cannot
experience deflection at the left-hand support.
x w(L)=0. The beam is also pinned at the right-hand support.
x w''(0) =0. As for the cantilevered beam, this boundary condition says
that the beam is free to rotate and does not experience any torque. In
real life, there is usually a small torque due to friction between the beam
and its pin, but if the pin is well-greased, this torque may be ignored.
x w''(L)=0. In the same way, the beam does not experience and bending
moments on its right-hand attachment.
The Euler-Bernoulli beam equation is given as
݀ସ‫ݍ ݓ‬
‫ܫܧ‬ ൌ ሺ‫ݔ‬ሻ
݀‫ ݔ‬ସ ‫ܮ‬
On integrating the Euler-Bernoulli beam equation we get the derived
equations as follows.
ࢊ૜ ࢝ ࢗ ࢞૛ (1)
ࡱࡵ ൌ ‫ כ‬൅ ࢉ૚
ࢊ࢞૜ ࡸ ૛
ࢊ૛ ࢝ ࢗ ࢞૜ (2)
ࡱࡵ ൌ ‫ כ‬൅ ࢉ૚ ࢞ ൅ ࢉ૛
ࢊ࢞૛ ࡸ ૟
ࢊ࢝ ࢗ ࢞૝ ࢞૛ (3)
ࡱࡵ ൌ ‫כ‬ ൅ ࢉ૚ ൅ ࢉ૛ ࢞ ൅ ࢉ૜
ࢊ࢞ ࡸ ૛૝ ૛
ࢗ ࢞૞ ࢞૜ ࢞૛ (4)
ࡱࡵ ‫࢝ כ‬ሺ࢞ሻ ൌ ‫כ‬ ൅ ࢉ૚ ൅ ࢉ૛ ൅ ࢉ૜ ࢞ ൅ ࢉ૝
ࡸ ૚૛૙ ૟ ૛
Now, let us consider that the load acting on the beam is uniformly distributed.
So, to calculate the load at a particular point on the beam the function we use
is
‫ݍ‬
ሺ‫ݔ‬ሻ
‫ܮ‬
Where,
(x = 0.1 to L)
55
Applying the boundary conditions in the above equation (4), we get the values
of constants by using the MATLAB software as.
C1 = -71.4726, C2 = 0, C3= 1.6885, C4 = 0
After solving this in MATLAB we came up with a graph showing maximum

deformation at the center.
The value obtained is 5.3*10-7m, but the obtained value at the center of beam
in simulation of part-2 is 6.83*10-7m which is close. But if we observe more
closely the deflection obtained from hand calculation is lesser than the one
which we obtained from simulation.
They are many reasons for that, it might be loading because our
loading is very approximate taken for the convenience of calculation. While
the simulation environment might have more realistic loading. The other
reason might be the material (density). For our mathematical model we have
ignored all the groves and holes and took the beam as normal solid beam, but
where it has multiple grooves and holes more material (density) means less
deflection. With all these observations and results, we can say the simulation
performed is close to reality and are logical.
56
7 Results & Discussions
7.1 Stress Analysis Results
The ultimate purpose of any analysis is to allow the comparison of the
developed stresses, strains, and deflections with those that are allowed by the
design criteria. All structures, and components therefore, must obviously be
designed to have a capacity greater than what is expected to develop during
the structure's use to obviate failure. The stress that is calculated to develop in
a member is compared to the strength of the material from which the member
is made by calculating the ratio of the strength of the material to the calculated
stress. But, before the simulation, one must be able to figure out where the
maximum deflection occurs in a part provided you have the information of
boundary conditions. For example, if I see a cantilever beam with a load acting
on it downwards, I can figure out that the maximum deflection would be at the
free end and there would be a reaction force present at the fixed end acting
opposite to the load. Similarly, we know the boundary conditions and loads
acting on each part at every time step. Therefore, we have some idea where
maximum deflection is going to be.
The Autodesk Inventor gives the stress distribution or deformation over the
part at a particular instant. If I want the stress distribution at a particular time
step, say 4.3rd second, I can edit the study properties, change the time step and
re-simulate it. But, Inventor doesn’t give the behavior of stresses along with
time. That means, we cannot able to see a Time vs Stress plot in Inventor. So,
just to understand the behavior of the stresses in a cycle and to find out where
the stresses are maximum or minimum in the cycle, we have found out a way.
The output grapher in the Dynamic simulation module of Autodesk Inventor,
we can find the model tree of all the part, joints, stable parts, forces arranged
there. By clicking any of them gives the option to view its plot against time.
Here, we arrive at joints and select a joint. If we open that joint, we can find
sub columns of position, velocity, acceleration, imposed force and moment,
each having their plot against time throughout the cycle. We select the moment
graph. The reason behind selecting the moment graph is,
We know that,
57
‫ܧ ܯ‬
ൌ
‫ܫ‬ ߩ
Where ‫ ܯ‬is the moment,
‫ ܧ‬is the Young’s modulus of the material,

‫ܫ‬ is the Moment of Inertia and
ߩ is the density of the material
The stress denoted by σ(y), can be written as the following equation,

െ‫ܧ‬ሺ‫ݕ‬ሻ
ɐሺሻ ൌ
ߩ
ா ெ
The term replaced by gives
ఘ ூ
െ‫ܯ‬ሺ‫ݕ‬ሻ
ɐሺሻ ൌ
‫ܫ‬
In the above equation, the moment of inertia is a constant value as the
dimensions of the part remain same over the cycle. Therefore, the moment is
proportional to the stress. Moreover, if we consider any part in the robotic arm,
the maximum deflection is always at the closer to the joints i.e.; at the ends.
The moment of the rotating part creates the deflections at those particular areas.
Considering all these reasons, we can say that the Time vs Stress plot will be
like Time vs Moment plot. But, to validate our assumption, we have provided
with the following information. The Time vs Stress graph of Part 1 is shown
below figure 7.1. We have simulated the part 1 in all the 30 time steps and plot
the obtained maximum stress values against time steps. Both are shown below
figure 7.1. They are not exactly similar because the Time vs Stress plot is
generated by plotting 30 time steps against the respective stress values in
Matlab. But the graph generated by Inventor is finer and definitely has more
than 30 time steps. So, we can see more smooth curve than the one generated
by the Matlab. But, the time steps where maximum and minimum values of
58
stress occur and time steps where maximum and minimum moment acts are
the same.
Figure 7.1.1: The graph between Time (sec) vs Stress of Part 1.
This supports our assumption. As simulating each part for 30 times and
recording the stress values is a complicated process, we can use Time vs
Moment plots to check those particular time steps where stresses are maximum
and minimum. To check this again, we have done all the process explained
above for part 2 also. Those two plots are also shown to be similar, making our
point stronger. The comparison is shown in appendix. But, a different process
for Part 5 is applied. There are 8 suction cups on Part 5 and the forces acting
on those 8 cups is same. But change in the position of the sheet with time
changes the behavior of forces are moments on Part 5. But all those Force vs
Time graphs and Moment vs Time graphs for Part 5 have their peak and lowest
values at the same point for all the suction cups. We can find the maximum
and minimum time steps using those figures which are shown below figure
7.1.1.
59
Figure 7.1.2: The graphs between Time (sec) vs force and moment.
The figure 7.1.2 below shows the stress distribution of Part 1. The stress
as expected, is maximum at the top where it forms a joint with part 2. Part 2
has a shaft that rotates inside the holes provided in the top of part 1. The
movement of part 2 induce stresses on part 1 on the top. There are also stresses
in the remaining part of the body. Complete blue color doesn’t say that there
are no stresses present at all. The division of these stress ranges over the scale
can be edited using the color plot option in inventor. Reduce the maximum
value and more detail stress distribution is seen. The maximum stress on this
part is 2.101 KSI which is equal to 14.48 MPa.
60
Figure 7.1.3: The stress distribution plot of Part 1.
Similarly, as shown in the figure (load section) in the Loads section,

Part 2 is also connected with a part on both of its ends. The right end of Part 2
is connected with Part 3. Therefore, the moment on the Part 3 shaft which
makes it rotate, have its effect on deflection in Part 2. Also, Part 2 is connected
to Part 1. So, the maximum stress values are developed at both the ends of Part
2. In the same way, Part 3 is also connected with a part at each end. The kind
of forces that are acting on Part 3 are like part 2. The most important
observation form the stress distribution pictures down here is that the critical
areas (where stresses are maximum) are found out to be at the ends of both Part
2 and Part 3, not anywhere in the middle. As we have discussed earlier, we
expect the critical areas to be at the ends due to loading at the ends. If by any
chance the critical area is spotted at the middle of either Part 2 or Part 3, it can
be bad. It is because, if the stresses are seen maximum at those areas where
loading is not more correspondingly, this means that the stress distribution of
that particular geometry is not proper and tend to fail soon. Both the stress
61
distribution images for Part 2 and Part 3 are shown below figures 7.1.3 and
7.1.4. The maximum stress found in Part 2 is 2.425 KSI which is equal to 16.71
MPa. The maximum stress found in Part 3 is 4.607 KSI which is equal to 31.76
MPa.
Figure 7.1.4: The Stress distribution of Part 2.
62
Part 4 and Part 5 both constitutes the end effector. Part 4 shown below figure
7.1.5 is a smaller part when compared to the size of remaining parts. It is
suspended at the end of part 3 and rotates inside the holes on the right end of
Part 3. It has 2 jobs to do. First one is to carry the weight if the end effector
and move itself in correspondence with Part 2 and Part 3. The second job is to
aid the rotation of end effector on its own axis. Naturally, as the load is acting
on its center, the deformation is expected to be more at the center. The sides
experience less deformation when compared to centers. The maximum stress
on this part shown below figure 7.1.5 is found out to be 0.235 KSI which is
equal to 1.67 MPa.
63
Part 5 is the backbone of end effector. The only duty it must hold the
sheet. So, the maximum effect created by the pneumatic cylinders and the sheet
is seen on this part. There are a lot of pretreated stresses in this part as we
assume that the pneumatic cylinders are welded to this part (just to lock the
DOFs of the cylinders). But those cylinders are assembled using screws and
nuts. If observed clearly, we can see that the end effector is provided with the
bases with holes and thread rods just for the cylinders to sit in. Due to the
welded joints created on this part, the stress values that are generated by this
simulation might be a bit more than the values. The stress distribution of part
5 is shown below figure 7.1.6. The maximum stress observed is 15.63 KSI
which is equal to 107.13 MPa.
Now, we have the stress distribution data for all the 5 major parts here but
to know whether these stresses are safe or not, we must check them. To
perform this task, we take the help of the factor of safety concept. But before
we get into this, the values of stresses shown along with these figures are at
one particular time step. But, the whole process is for 7.5 seconds. They also
64
vary with the time and this point has already been explained along with Time
vs Moment plot earlier in this section. To know where maximum and minimum
stress occurs in the whole cycle, we take the help of Time vs Moment plots of
all the parts and take out the maximum and minimum time steps in those plots.
Those values are tabulated below in table 7.1.7.
65
Table 7.1.1: The Table of maximum and minimum stresses.
PART MAX. TIME STEP MIN.STRESS TIME STEP

STRESS AT WHICH AT WHICH
MAX. MIN. STRESS
STRESS OCCURS
OCCURS
1 15.72 MPa 0.54 sec 14.14 MPa 0.136 sec
2 20.408 MPa 0.54 sec 4.52 MPa 0.136 sec
3 4.136 MPa 1.09 sec 1.31 MPa 0 sec
4 6.51 MPa 1.09 0.387 MPa 0 sec
5 15.92 MPa 1.22 sec 4.49 MPa 0 sec
The Factor of Safety is a term describing the load carrying capacity

of a system beyond the expected or actual loads. Essentially, the factor of
safety is how much stronger the system is than it usually needs to be for an
intended load. Safety factors are often calculated using detailed analysis
because comprehensive testing is impractical on many projects, such as
bridges and buildings, but the structure's ability to carry load must be
determined to a reasonable accuracy. Many systems are purposefully built
much stronger than needed for normal usage to allow for emergency situations,
unexpected loads, misuse, or degradation. Factor of safety is the ratio of Yield
stress and Working stress. It gives a value which gives the load capacity of the
part. If Factor of Safety is 1, this means the design can carry the load exactly
that is required to carry. If the value is 2, this means it can carry twice the
required load. Let us see the factor of safety values generated by Inventor for
all the parts.
As we see that the Part 5 is not as strong as expected. The safety
factor it has permits it only to carry the load of the sheet, which is good for this
case but we expect the design to be stronger than this. There are many ways to
make this stronger. We can either optimize it and redesign it accordingly or
66
can try a different material. The Part 5 is the end most part of the robotic arm.
Part 5 showed less load bearing capacity with aluminum but, we have tried
changing the material to CFRP (Carbon Fiber Reinforced Polymer). The
reason for this is it is stronger and lighter material than aluminum. We cannot
assign a material with more density. As density increases, weight increases and
increase in weight of Part 5 might have bad effects on other arms. Our
experiment worked and the part 5 has increased its capacity. The only problem
with CFRP is it is relatively expensive. But again, selection of materials is a
completely different study. We limit our study to the usage requirements of the
industry.
Figure 7.1.8: The FOS of Part 5.
These results that are seen are good at this instant. The model can take
the weight of the sheet and if wanted, it can take more. But, this cannot be
taken as the final word. The process in which the robotic arm is going to be
used is a cyclic process. The same kind of loading is applied on the robotic arm
repeatedly for a long period. We must also check for the approximate number
67
of cycles it can work before it fails. This leads us to the Fatigue Analysis of
the Robotic arm.
7.2 Convergence
In the process of meshing the geometry and solving for the stress analysis, the
stress results are needed to be found out more accurately. To achieve this, the
geometry is meshed normally first and simulated. Then, the areas with more
stress concentration is chosen and are meshed finely when compared to the
previous mesh. Once the results are obtained, the areas with more stress
concentration are meshed more finely. Here, for every step of the solution, the
number of elements in the mesh increase. As mesh settings change, the results
of a Stress Analysis may change and the accuracy of the stress results should
be understood. Performing a convergence study refines the mesh and reduces
the size of the elements, which will theoretically increase the accuracy of the
next iteration of results. As mesh elements decrease in size but increase in
quantity, the computational requirements to solve a given model increase. As
mesh elements decrease in size, they reach a point of diminishing returns on
the level of accuracy compared to the computational overhead and time
required to compute the result. This means that a simulation requires more time
to compute the results, but the result may change by an insignificant value. The
solution is said to be converging if the stress value falls around same number
as the number of time steps increases. The maximum stresses for part 2 are
plot against the solution steps in the graph that is shown below. There is a
significant increase in the number of elements for the increase of each solution
step. The graph is plot using Matlab. The plot can be seen in figure 7.2.1. As
seen in that plot, the stresses vary with increase in time steps (X-axis). As
mentioned earlier, the number of mesh elements increases with time steps.
That means, the peak most point in the curve has more number of meshed
elements and therefore can be taken as the result closer to accuracy. As seen in
the convergence plot, the last two stress values in the curve are very close to
each other making the curve linear onwards. This is the proof that the solution
converges and this stress value can be taken as the closest value to the stresses
in reality. The values of stresses for each time step and the number of meshed
elements for each time step are tabulated below in the table 7.2.1.
68
Figure 7.2.1: The Convergence Plot for Part 2.
Table 7.2.1: The Table of Number of Mesh Elements and Time Steps.
Time Steps Number of Maximum
Mesh Von Mises
Elements Stress KSI
1 1702 1.195
2 2051 1.592
3 4569 1.691
4 8988 3.139
5 22689 4.717
6 82850 5.07
7 271968 5.021
69
7.3 Fatigue Analysis
Fatigue is the weakening of a material caused by repeatedly applied
loads. It is the progressive and localized structural damage that occurs when a
material is subjected to cyclic loading. The nominal maximum stress values
that cause such damage may be much less than the strength of the material
typically quoted as the ultimate tensile stress limit, or the yield stress limit.
When cyclic stresses are applied to a material, even though the stresses do not
cause plastic deformation, the material may fail due to fatigue. Fatigue failure
is typically modeled by decomposing cyclic stresses into mean and alternating
components. Mean stress is the time average of the principal stress. The
definition of alternating stress varies between different sources. It is either
defined as the difference between the minimum and the maximum stress, or
the difference between the mean and maximum stress. Engineers try to design
mechanisms whose parts are subjected to a single type (bending, axial, or
torsional) of cyclic stress because this more closely matches experiments used
to characterize fatigue failure in different materials. For this, we need to have
an approximate period where we will be able to assess that the material or
design fails after these many number of cycles.
Fatigue failures are typically characterized as either low-cycle (1,000
cycles). The threshold value dividing low- and high-cycle fatigue is somewhat
arbitrary, but is generally based on the raw materials behaviour at the
microstructural level in response to the applied stresses. Lowcycle failures
typically involve significant plastic deformation. Most metals with a body
centered cubic crystal structure have a characteristic response to cyclic
stresses. These materials have a threshold stress limit below which fatigue
cracks will not initiate. This threshold stress value is often referred to as the
endurance limit. In steels, the life associated with this behaviour is generally
accepted to be 2x106 cycles. In other words, if a given stress state does not
induce a fatigue failure within the first 2x106 cycles, future failure of the
component is considered unlikely. For spring applications, a more realistic
threshold life value would be 2x107 cycles. Metals with a face center cubic
crystal structure (e.g. aluminium, austenitic stainless steels, copper, etc.) do
not typically have an endurance limit. For these materials, fatigue life
continues to increase as stress levels decrease; however, a threshold limit is not
typically reached below which infinite life can be expected. The behaviour of
70
these materials explained above can be exhibited with the help of S-N diagram,
also known as a Wohler curve. This is a graph of the magnitude of a cyclic
stress (S) against the logarithmic scale of cycles to failure (N). The progression
of the S-N curve can be influenced by many factors such as corrosion,
temperature, residual stresses, and the presence of notches. The Goodman-Line
is a method used to estimate the influence of the mean stress on the fatigue
strength[14]. Fatigue data on 6061-T6 aluminium were collected from the
literature (Structural Alloys Handbook, 1989; Alcoa Structural Handbook,
1960 to construct a design fatigue curve. The low cycle data (l03 cycles) are all
from axial strain controlled tests. Note that the data from the strain controlled
tests fall right in with the rest of the data at greater than 103 cycles. Some of
the high cycle data are from rotating beam tests. A fit to the 6061-T6 fatigue
data was developed using the equation below.
‫ܧ‬ ͳͲͲ
ܵൌ ൬ ൰൅‫ܤ‬
Ͷξܰ ͳͲͲ െ ‫ܣ‬
Where E = elastic modulus
N = Number of cycles to failure
S = Strain Amplitude times elastic modulus
and A and B are constants that are selected to make the equation fit the data.
Note that S is treated like a stress based on the assumption of elastic behaviour.
However, it does not represent a real stress when the elastic range is exceeded.
As a first approximation, the fatigue properties can be estimated by taking A
as the percentage reduction of area in a tensile test, RA, and B as the endurance
limit Se. It should be noted that aluminium does not exhibit an endurance limit.
The fit to the data is
ଵସସ଻ଽ
ܵ ൌ ൅ ͻ͸ǤͷMPa
ξே
The curve is shown with the available data from tests without any mean
stress in Figure 7.3.1.
71
Figure 7.3.1: Aluminium 6061 Fatigue Data from Experiments[15].
Goodman relation is mathematically represented as

ߪ௔ ߪெ
൅ ൌͳ
ߪ௘ᇱ ߪ௨
Where ߪ௔ is alternating stress,
ߪெ is mean stress,
ߪ௨ is ultimate stress and

ߪ௘ᇱ is effective alternating stress.
Let us find the number of cycles to the failure one of the parts in the robotic
arm, for example Part 2. From the table (7.1.1), we have
72
Maximum stress, ߪ௠௜௡ = 4.521 MPa
ߪ௠௔௫ = 20.408 MPa

From these values, we now have,
Stress range ߂ߪ= ߪ௠௔௫ െ ߪ௠௜௡

௱ఙ
alternating stress ߪ௔ =
ଶ
ఙ೘ೌೣ ାఙ೘೔೙
mean stress ߪெ =
ଶ
The material used for this part is Aluminum. The data needed for the
calculation of Number of Cycles is shown below.
The Ultimate tensile strength ܵ௨௧ = 446 MPa
The Material Strength at 103 cycles ܵ௠ = 0.95ൈܵ‫ݐݑ‬
Since the load acting on the part is bending load the constant value is taken as
0.95. If it is an axial loading, the constant value is taken as 0.75. Aluminum
has no endurance limit. Some structural metals such as aluminum and copper,
do not have a distinct limit and will eventually fail even from small stress
amplitudes. In these cases, several cycles (usually 107) is chosen to represent
the fatigue life of the material.
Endurance Limit for this part ܵ௘ = ‫ܭ‬௔ ‫ܭ‬௕ ‫ܭ‬௖ ‫ܭ‬ௗ ‫ܭ‬௘ ‫ܭ‬௙ ܵ௘ᇱ
Where ܵ௘ᇱ is the experimental value of endurance limit R. R. Moore [16]

experiment.
ܵ௘ᇱ = ͲǤͷൈܵ௨௧ ; ሺܵ௨௧ ൑ ͳͶͲͲ‫ܽܲܯ‬ሻ
ܵ௘ᇱ = ͹ͲͲ‫ ;ܽܲܯ‬ሺܵ௨௧ ൐ ͳͶͲͲ‫ܽܲܯ‬ሻ
x ‫ܭ‬௔ is the surface factor

‫ܭ‬௔ =ܽሺܵ௨௧ ሻ௕ where ܽ is the surface finish factor [17] which is 4.51
ܾ is the exponent which is -0.265
73
‫ܭ‬௔ =0.023
x ‫ܭ‬௕ is the size factor

‫ܭ‬௕ = -1.51ൈ݀ ି଴Ǥଵହ଻ ; (51 ൏ ݀ ൑ ʹͷͶ݉݉) [17]
݀ is generally referred to the diameter of the body. Since non-circular bodies
have no particular diameter, we must calculate an effective diameter ݀௘ to
formulate the above equation. Considering our part as a rectangular part, the
effective diameter is calculated as shown here.
݀௘ =0.808ൈሺ݄ൈܾሻଵȀଶ [17]
Where ݄ = 0.153 mm and ܾ = 0.08 mm
‫ܭ‬௕ = 7.45ൈͳͲିସ
x ‫ܭ‬௖ is the Loading Factor

ͳܾ݁݊݀݅݊݃
‫ܭ‬௖ =൝ ͲǤͺͷܽ‫ ݈ܽ݅ݔ‬ൡ
ͲǤͷͻ‫݊݋݅ݏݎ݋ݐ‬
x ‫ܭ‬ௗ is the Temperature factor

்ܵ
‫ܭ‬ௗ ൌ
ܵோ்
Where ்ܵ is the tensile stress at the T temperature and

ܵோ் is the tensile stress at the room temperature.
Since, our machine is placed at the working conditions in the room
temperature, ‫ܭ‬ௗ ൌ ͳ
74
x ‫ܭ‬௘ is the Reliability Factor; ‫ܭ‬௘ ൌ ͳ
x ‫ܭ‬௙ is the Miscellaneous-Effects Factor.

Though the factor ‫ܭ‬௙ is intended to account for the reduction in endurance
limit due to all other effects, it is intended as a reminder that these must be
accounted for. The actual values of ‫ܭ‬௙ are not available.
‫ܭ‬௙ ൌ ͳ
Now, the endurance limit ܵ௘ = 3.825ൈͳͲଷ ‫ܽܲܯ‬
If we take ܵ௡ as the stress at ݊ number of cycles, the equation to calculate ܵ௡
will be
ܵ௡ ൌ ‫ܣ‬ൈܰ ஻
Where‫Ͳͳ = ܣ‬ሺ௅௢௚ሺௌ೘ ሻିଷ௕ሻ
ଵ ௌ
‫=ܤ‬ቀ ቁ ൈቀ ೘ ቁ
௓ ௌ ೐
Z= ሺܰଵ ሻ െ ሺܰଶ ሻ
The boundary conditions for the materials without endurance limit (example
Aluminum)
ܵ௡ ൌ ܵ௠ ܽ‫ ܰݐ‬ൌ ܰଵ ൌ ͳͲଷ
ܵ௡ ൌ ܵ௙ ܽ‫ ܰݐ‬ൌ ܰଶ ൌ ͷൈͳͲ଼
Now, considering the Goodman equation,
ߪ௔ ߪெ
൅ ൌͳ
ߪ௘ᇱ ߪ௨
ߪ௔ ߪெ
ൌ ͳ െ
ߪ௘ᇱ ߪ௨
75
ߪ௔
ߪ௘ᇱ ൌ ቌ ߪெ ቍ
ቀͳ െ ቁ
ߪ௨
But we have another equation ߪ௘ᇱ ൌ ܵ݊ ൌ ‫ܣ‬ൈܰ‫ܤ‬

భ
ఙ೐ᇲ ್
N=ቀ ቁ
஺
Where N is the number of cycles which in case the calculation 4.3ൈ(1017)
7.4 Stress-Cycle (S-N Diagram)

In high-cycle fatigue situations, materials performance is commonly
characterized by an S-N curve. This is a graph of the magnitude of an
Alternating Stress Amplitude (ߪ௘ᇱ ) against the logarithmic scale of cycles to
failure (N). We have plot them on the S-N diagram using the above data in
Matlab. The plot obtained is shown below figure 7.3. Also the table 7.3.1
shows the number of cycles to failure for each part in the robotic arm.
Table 7.3.1: The Table of Number of Cycles to Failure for each part.
Part Number of
Cycles to
Failure
1 5.8 x 1018
2 4.3 x 1017
3 3.1 x 1018
4 1.29 x 1018
5 6.28 x 1017
76
Figure 7.4.2: The graph between Time (sec) vs Stress (Mpa).
As seen, the plot above, the blue colored curve shows the stresses over
the time. The red line is the line where we can see the stress amplitude or
fatigue strength of the part which is calculated from the simulation.
Theoretically speaking, the point where these 2 curves meet is the failure point
of the material. But, as we can see that they do not meet, there is no scope of
failure for this part theoretically. The life of the part ends before 103 cycles, it
is the Low Cycle Fatigue. If it lasts till 108 cycles, it is the High Cycle Fatigue.
If the part lasts or more than 108 cycles where the stress curve and endurance
limit or fatigue strength become parallel to each other, then it has infinite life.
This shows that this part is going to withstand for a good period of time.
77
8 Summary &Conclusions
8.1 Validation
For any design to be judged as a good design, it must be proven that it
is useful for the purpose it is made for. In our case, we have analysed our model
by creating the simulation environment which is mostly close to the actual
working conditions and we have obtained some results. Now to check whether
these results are satisfactory, we must take the Factor of Safety concept. Factor
of safety is generally the ratio of Yield Stress of the design to the allowable
stress. But Inventor has an option to set the ratio between Ultimate Tensile
Stress to the allowable stress. A yield strength or yield point is the material
property defined as the stress at which a material begins to deform plastically.
Prior to the yield point the material will deform elastically and will return to
its original shape when the applied stress is removed. Ultimate tensile strength
is the capacity of a material or structure to withstand loads tending to elongate,
as opposed to compressive strength, which withstands loads tending to reduce
size. In other words, tensile strength resists tension (being pulled apart),
whereas compressive strength resists compression (being pushed together).
Ultimate tensile strength is measured by the maximum stress that a material
can withstand while being stretched or pulled before breaking. If we observe
the Stress-Strain curve, we can see that the Ultimate tensile strength always
comes after the yield point. This point serves a very important observation.
78
Figure 8.1.1: The Factor of Safety for Part 2.
If the Factor of safety of the parts in robotic arm is more than 3 when
the ratio is yield stress to working stress, it will be much more than 3 if the
ratio is Ultimate tensile strength to working stress. This means, the robotic arm
far away from failure even its weight is increased by unavoidable accessories
like servo motors, protective covers and probably a different and denser end
effector. Our study shows that the robotic arm is not only strong enough for
the required use but also strong enough to work for a long time without failure.
Here, the factor of safety simulation figure 8.1 is shown above. All the parts of
this robotic arm are having the minimum factor of safety more than 3 in those
areas where the stresses are more. The table 8.1 below showing the factors of
safety along with their stresses and materials assigned.
79
Table 8.1.1: The Table of Factors of Safety of all Parts
Part Material Minimum Safety

Factor
1 Aluminium 6061, 3.89
welded
welded
welded
welded
5 CFRP 3.38
8.2 Conclusion
The main reason to make the choice of designing a robotic arm, firstly
is the transport of the sheet from the stack to the shearing machine. Second,
there is a space constraint we have in the industry. The stack of the sheets and
the shearing machine are close to each other. They are so close that any other
kind of transport such as a linear robot (conveyor) cannot be used. That leaves
us with the only choice to build a robotic arm which can work in small and
congested environments. The figure below shows the location and position of
stack and shearing machine and we must design a robot arm that fits in this
environment. As we observe the document, we can see that we have tested the
robotic arm in all possible ways such as its load bearing capacity, motion
ranges, joints etc. The motion demonstration of the robotic arm can be seen
through the figures given in the appendix. The main task was to design a
machine which fits in the environment and serves the purpose which has been
achieved. But this is not the ultimate design. We tried to cover all the aspects
of design and structural analysis in our work. But there is still a large chance
of upgrading the machine.
80
As far as the current study is concerned, we are successful in
determining the areas in which the stresses occur the most and able to show
that those stresses are in limit also do not cause any considerable damage to
the robot arm. The stress analysis that was performed was the non-linear stress
analysis and all the geometric non-linearity is considered in the simulation.
This shows that the stresses and deformation obtained are very close to the
reality. The duration of the half cycle of the robotic arm in our simulation is
7.5 seconds, it means the sheet will be picked, transported and placed at its
destination in 7.5 seconds which is a quick process This speed might not be
compatible in the real-life use in the industry. We might have to slow down
the process a bit. This 7.5 seconds’ time limit isn’t put out of blue. After many
adjustments, trials, redesigns we have come to this time from 18 seconds to 7.5
seconds. Our motive to simulate it in such speed is that, if this machine can
with stand motion loads in high speeds and still be safe, it can be a good
machine at lower speeds. The speeds can always be adjusted by the user.
Figure 8.2.1: The position of Shearing Machine in Industry.
81
8.3 Future Works
Although a good effort is put in building and analysing this system,
this cannot be taken for granted. There are many other parameters on which
the quality of a model is dependent on. The easiest way to design and analyse
a robotic arm is described and explained in this document. We cannot say it is
perfect. In fact, no design is perfect. It is all about how close you can get to
perfection. This model is needed to be optimized. We must have a detail
explanation on what effect could take place on what area of the robotic arm if
a parameter is changed especially design parameters like volume, thickness,
material. Also, the cost study is needed to be performed in a more detail way.
After all, a good machine is also judged on how affordable it can be.
There is always more room for innovation in any study. During this
project, we came across many modern innovative automation techniques
today’s world has developed. If given a chance to do this project again, we are
very much interested in applying those techniques. One such technique is
application of fluidic muscles. A fluidic muscle is artificial version of human
muscle. Not only this, many other techniques are applicable for this kind of
machine. With all these additions, the machine gets upgraded into a super
machine with ultimate application of automation.
82
References
[1] Minoru Baba, “Pick and Place for automatic test handler.” .
[2] Yang Liu and Michael Yu Yang, “Optimal Design of Remote Centre
Compiance Devices of Rotational Symmetry.” .
[3] “Robot end effector,” Wikipedia, the free encyclopedia. 27-Jun-2016.
[4] M. M. B. S. Pachaiyyapan and T.Sridhar, “Design and Analysis of an
Articulated Robotic Arm for various Industrial Applications,” 2015.
[5] FANUC America Corporation, Automated Bottle Pallet Unloader with
FANUC Pick & Place Robot – Clear Automation. 2016.
[6] “Understanding Robot Specifications 101.” [Online]. Available:
https://www.robots.com/blog/viewing/understanding-robot-
specifications-101. [Accessed: 30-Sep-2016].
[7] “Inventor Forum.” .
[8] majentauk, Simulation Tips - Autodesk Inventor 2014. 2014.
[9] “Materials for Robot building an introduction.” .
[10] “Ashby charts.” Mike Ashby.
[11] “Aerospacs Specification Metals.” .
[12] Simulation Tips - Autodesk Inventor. .
[13] “Beam Boundary conditions.” .
[14] R. Stone, Fatigue Life Estimates using Goodman Diagrams. .
[15] Yahr, G.T., “36 Materials Science; Aluminium; Fatigue; Research
Reactors; Presure Vessels; Welded Joints; Neutron Sources 360103;
Mechanical Properties.” Office of Scientific and Technical Information,
01-Jun-1993.
[16] R.R. Moore, “Rotating Beam Fatigue Testing System.”
[17] K. N. R.Budnyas, Shigley’s Mechanical Engineering Design, Ninth.
Shigley, 2003.
83
Appendix
Link to the Motion Demonstration of the Robotic Arm
https://www.youtube.com/watch?v=W3ImRT73SVk&feature=youtu.be
A. External Forces Acting on Parts
Figure A.1. The forces and moments of Part 1
84
85
86
B. Deformations on Parts
Figure B.1. The deformation of Part 1
87
Figure B.3. The deformation of Part
88
89
C. Moment Vs Time Graphs
Graph C.1. The moment vs time plot for Part 2-3
Graph C.2. The moment vs time plot for Part 3-4
90
Graph C.3. The moment vs time plot of Part 4-5
91
D. The Factor of Safety on Parts
Figure D.1. The Factor of Safety of Part 1
92
93
E. Von-Misses Stress of Part 5 with CFRP material
Figure E.1. The Von-Misses Stress of Part-5 with CFRP material
94
F. Matlab Code for calculating the Number of Cycles
Sut=446e+6; %Ultimate Tensile Stress
Sm=0.95*Sut; %Material Strength at 10^3 cycles
Se1=0.5*Sut; %Endurance Limit
h=153; %Height of the cross section-meters

br=80; %Breadth of the cross section-meters
de=0.808*((h*br)^0.5); %Effective Diameter
A=4.51;
Bb=-0.265;
Ka=A.*(Sut.^Bb); %Surface Factor
Kb=1.51*(de^(-0.157))*0.001; %Size Factor
Kc=1; %Loading Factor
Kd=1; %Temperature Factor
Ke=1; %Reliability Factor
Kf=1; %Miscellaneous Factor
Se=Ka*Kb*Kc*Kd*Ke*Kf*Se1;
N1=10e+3;
N2=5*(10e+8);
Z=log(N1)-log(N2);
b=(1/Z)*log(Sm/Se);
a=10^(log(Sm)-(3*b));
Smin=4.49e+6; %Minimum Stress in the cycle on the part

Smax=15.92e+6; %Maximum Stress in the cycle on the part
Srange=Smax-Smin; %Stress Range
Salter=Srange/2; %Alternating Stress
Smean=(Smax+Smin)/2; %Mean Stress
Syeild=296e+6; %Yeild Stress
P=Smean/Sut;
Q=(1-P);
Seffective=Salter/(Q); %Effective Alternating Stress at
Faliure
N=(Seffective/a)^(1/b); %Number Of Cycles for Failure
%Matlac code for S-N Curve

Sut=(50:50:650)*10^6; %Ultimate Tensile Stress
Sm=0.95*Sut; %Material Strength at 10^3 cycles
Se1=0.5*Sut; %Endurance Limit
h=153; %Height of the cross section-meters

br=80; %Breadth of the cross section-meters
de=0.808*((h*br)^0.5); %Effective Diameter
A=4.51;
95
Bb=-0.265;
Ka=A.*(Sut.^Bb); %Surface Factor
Kb=1.51*(de.^(-0.157))*0.001; %Size Factor
Kc=1; %Loading Factor
Kd=1; %Temperature Factor
Ke=1; %Reliability Factor
Kf=1; %Miscellaneous Factor
F=Ka.*Kb;
D=Kc*Kd*Ke*Kf*Se1;
Se=F.*D;
N1=10e+3;
N2=5*(10e+8);
Z=log(N1)-log(N2);
b=(1/Z)*log(Sm./Se);
a=10.^(log(Sm)-(3.*b));
Smin=4521580; %Minimum Stress in the cycle on the part

Smax=20408400; %Maximum Stress in the cycle on the part
Srange=Smax-Smin; %Stress Range
Salter=Srange/2; %Alternating Stress
Smean=(Smax+Smin)/2; %Mean Stress
Syeild=296e+6; %Yeild Stress
P=Smean./Sut;
Q=(1-P);
Qq=1./Q;
Seffective=Salter*Qq; %Effective Alternating Stress at Faliure
M=(Seffective./a);
B=1./b;
N=M.^(B); %Number Of Cycles for Failure
plot(N,Seffective,'b');
hold on
C=(0:1:9)*10e+16;
E=[7.9e+6 7.9e+6 7.9e+6 7.9e+6 7.9e+6 7.9e+6 7.9e+6 7.9e+6
7.9e+6 7.9e+6];
plot(C,E,'r');
xlabel('Time Sec');
ylabel('Stress MPa');
title('S-N Diagram');
legend('S-N Curve','Stress Amplitude');
title('S-N Curve for Part 2');
grid minor
96
97
School of Engineering, Department of Mechanical Engineering Telephone: +46 455-38 50 00
Blekinge Institute of Technology E-mail: [email protected]
SE-371 79 Karlskrona, SWEDEN

FULLTEXT02

Uploaded by

Copyright:

Available Formats

FULLTEXT02

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FULLTEXT02

Uploaded by

Copyright:

Available Formats

Master's Degree Thesis

Design and Structural Analysis of

Gurudu Rishank Reddy

Department of Mechanical Engineering

Supervisors: R.V Jhansi Rao, Signode India Limited (SIG)

Department of Mechanical Engineering

Karlskrona, October 2016

CAD Computer Aided Design

CFRP Carbon Fibre Reinforced Plastic

DOF Degrees of Freedom

FANUC Fuji Automated Numerical Control

FEA Finite Element Method

FOS Factor of Safety

KSI Kilo pound per Square Inch

SIG Signode India Limited

RCC Remote Censor Control

• A fixed bed to which one blade is attached.

Shearing operation is generally conducted manually, but it can be

2.1 Project Statement

2.2 Background Study

2.3 Research Problem

2.6 Research Questions

2.7 Preliminary Discussion

2.7.1 Articulated Arm Robots

Articulated arm robots are generally used to perform risky,

2.7.2 End Effector of the Robot

2.8 Related Works

3.1 Mechanical Design

Figure 3.2.2: The shaft and Key of Oldham Coupling.

The dimensions of the shaft are given below.

Length of the part (L)= 0.45 meters

Length of the part (L)= 0.775 meters

Figure 3.4.1: The CAD design of Part 3.

3.5 Part 4 & Part 5

Length of Part 4 (L)= 0.266 meters

Figure 3.5.1: The CAD design of Part 4 and Part 5.

Figure 3.6.1: The CAD design of pneumatic Cylinder.

In the below figure 3.7.1, we can see 3 parts. The base is

W is the load acting on the shaft.

So, by considering the above dimensions the torque is to be calibrated.

Therefore, the torque is T= W * R

Torque of the solid shaft is = 31.614 Nm.

Here, as we know the radius of the shaft is r = 0.12meters

Therefore, Induced shear is = 31.614 * 16 / 3.14 *(0.24)3

Induced shear is = 11.65 MPa

Allowable shear = 0.18 * 310

Therefore, the factor of safety is = allowable shear / Induced shear

The Factor of safety of a solid shaft is = 4.75

This is the joint that is formed by the combination of Part 1

3.7.3 Joint 3 (Elbow)

3.7.4 Joint 4 (Wrist)

Wrist is the area which balances the operation. In this operation,

Figure 3.7.4: The CAD design Joint 4 (Wrist).

3.7.5 Joint 5 (End Effector)

Joint 5 is the most important and delicate part of this whole

Figure 3.7.5: The CAD design of Joint 5 (End Effector).

3.8 Dynamic Behaviour of Robotic Arm

Figure 3.8.1:The dynamic behaviour of Pneumatic Cylinder.

Figure 3.8.2: The Prototype of Robotic Arm in three sections.

Figure 3.8.3: The Top view of Robotic Arm.