UNIVERSITY OF MISKOLC
FACULTY OF MECHANICAL ENGINEERING AND INFORMATION SCIENCE
MASTER IN MECHANICAL ENGINEERING
CAD/CAM SPECIALIZATION
MODELING AND OFFLINE SIMULATION OF THERMAL SPRAY
COATING PROCESS
GRADUATION WORK PRIOR TO OBTAIN THE DEGREE OF
MASTER IN MECHANICAL ENGINEERING
AUTHOR:
NOUI YOUCEF RAMZI
PROJECT LEADER:
DR. HEGEDŰS GYÖRGY
MISKOLC-HUNGARY
2016
CONTENTS
1.
ABSTRACT .......................................................................................................................... 3
2.
INTRODUCTION .................................................................................................................. 4
3.
STATE OF ART IN THERMAL SPRAYING.............................................................................. 5
4.
5.
3.1.
Thermal spray modeling .............................................................................................. 5
3.2.
Operating parameters in thermal spraying ................................................................... 6
3.3.
Relative velocity torch/substrate ................................................................................. 7
3.4.
Spray distance .............................................................................................................. 8
3.5.
Projection angle ........................................................................................................... 9
3.6.
Offset path ................................................................................................................. 10
MODELING, ANALYSIS AND CONTROL OF ROBOT MANIPULATOR.................................. 11
4.1.
System description ..................................................................................................... 11
4.2.
Robot mechanics........................................................................................................ 12
4.3.
The workspace ........................................................................................................... 12
4.4.
Axis data .................................................................................................................... 14
4.5.
Main programing methods......................................................................................... 14
4.6.
Modeling .................................................................................................................... 14
4.7.
Geometric description................................................................................................ 15
4.8.
Robotic CAD systems ............................................................................................... 18
4.9.
Control of the robot ................................................................................................... 19
4.9.1.
Program by teaching on-line.............................................................................. 20
4.9.2.
Offline programming .......................................................................................... 20
4.9.3.
Programming by robotic language .................................................................... 21
4.9.4.
Graphical programming .................................................................................... 21
4.9.5.
Programming by self-learning ........................................................................... 22
EXPERIMENTAL EVALUATION OF PROCESS PARAMETERS AND MODEL VERIFICATION . 23
5.1.
Characterization of Deposit Thickness ...................................................................... 23
5.2.
Modeling of the singular strand of the deposit .......................................................... 23
5.3.
Mathematical model of the singular spray ................................................................ 24
5.4.
Mathematical description of the singular spray profile ............................................. 24
5.5.
Experimental protocols .............................................................................................. 25
5.6.
Acquisition of the CAD file....................................................................................... 26
5.7.
Selection of the thermal spray operating parameters ................................................. 30
5.8.
Trajectory generating ................................................................................................. 31
1
6. INTRODUCTION TO OFFLINE PROGRAMMING AND COATING SIMULATION WITH
ROBCAD SOFTWARE ............................................................................................................... 32
6.1.
7.
Spray robot programming in RobCAD ..................................................................... 32
6.1.1.
Workcell layout .................................................................................................. 32
6.1.2.
Path editor .......................................................................................................... 35
6.1.3.
Calibration and the adjustment of the virtual RobCAD work cell. .................... 36
6.2.
Application of the RobCAD/Paint software for coating thickness simulation .......... 37
6.3.
Summary of the simulation of coating deposition for Turbine blade ........................ 38
CONCLUSION .................................................................................................................... 39
REFERENCES ............................................................................................................................ 40
2
1. ABSTRACT
In manufacturing companies today, there is strong competition mainly driven by the axis of
technological development to find new manufacturing procedures to improve the quality/price
imposed by increased competitiveness. This leads to introduce the concept of flexible
manufacturing systems, through the use of robots and robotic sites in chain manufacturing.
Among the advantages often cited in favour of the introduction of robots are decreased labour
costs, increased precision and productivity, increased flexibility compared with specialized
machines, and more humane working conditions as dull, repetitive, or hazardous jobs are
performed by robots.
An official definition of such a robot comes from the Robot Institute of America (RIA): A robot
is a reprogrammable multifunctional manipulator designed to move material, parts, tools, or
specialized devices through variable programmed motions for the performance of a variety of
tasks.
The robot, as we have defined it, was born out of the marriage of two earlier technologies:
teleoperators and numerically controlled milling machines. Teleoperators, or master-slave
devices, were developed during the Second World War to handle radioactive materials.
Computer numerical control (CNC) was developed because of the high precision required in
the machining of certain items, such as components of high performance aircraft.
The first robots essentially combined the mechanical linkages of the teleoperator with the
autonomy and programmability of CNC machines. The first successful applications of robot
manipulators generally involved some sort of material transfer, such as injection molding or
stamping, where the robot merely attends a press to unload and either transfer or stack the
finished parts. These first robots could be programmed to execute a sequence of movements,
such as moving to a location. A closing a gripper, moving to a location B, etc., but had no
external sensor capability. More complex applications, such as welding, grinding, deburring,
and assembly need not only more complex motion but also some form of external sensing such
as vision, tactile, or force-sensing, due to the increased interaction of the robot with its
environment. It should be pointed out that the important applications of robots are by no means
limited to those industrial tasks where the robot is directly replacing a human worker. There are
many other applications of robotics in areas where the use of humans is not advised or
undesirable. Among these are undersea and planetary exploration, satellite recuperation and
repair, the defusing of explosive devices, and work in radioactive environments. Finally,
prostheses, such as artificial limbs, are themselves robotic devices requiring methods of analysis
and design similar to those of industrial manipulators.
3
2. INTRODUCTION
Robots are increasingly used in thermal spraying to replace human operators not only for the
needs of coating quality guarantee but also for the protection against noxious spray operating
environments such as high temperature; loud noises; poisonous gases etc. In the general process
of thermal spraying, a torch gun is installed on the wrist (the last axis) of the robot, thus the
robot can take the torch to go through the entire desired surface of workpiece. High
reproducibility and flexibility of the robot provide an achievement of high quality and economic
coatings on the substrate. The most important advantage of the robot is its flexible movements;
it can achieve a projection on some complex, free-form workpieces such as turbine blades,
frying-pans, etc. When using the robot, the reliability and functional properties of the coating
can be improved. Achieving uniform coating deposit for free-form workpiece is still an
interesting research topic due to the complex geometry of surfaces. The use of accurate robot
programs is necessary to ensure coating properties that require low tolerances concerning
thickness, roughness, hardness. Those parameters shall be achieved by new methods to translate
the operating parameters of plasma arc spray into robot motion and generate paths on controlled
points by subdividing the trajectory with FEM nodes that all connected to each other will appear
as continuous pathway on a free form surface. The aim off this work is to generate a functional
robot pathway for complex motion in a virtual but realistic cells to avoids expensive trials
before implementation on site and apply modification and optimization even in ongoing
process.
4
3. STATE OF ART IN THERMAL SPRAYING
Some text need to be placed here about thermal spray technology!
3.1. Thermal spray modeling
The plasma jet operating principle lies in the transfer of energy from an electric arc discharge,
generated between cathode and anode, to the process gas flowing between them. Argon is
usually used as the primary process gas with the addition of secondary gases such as hydrogen,
helium or nitrogen [10]. The process gas is ionized by an electrical discharge sustained by DC
power [9]. The ionized gas creates high-pressure electrically neutral plasma which leaves the
nozzle at high speed and temperature. Coating powder is usually fed by injectors into the plasma
jet either from outside of the gun or directly in the diverging exit region of the nozzle (external
and internal powder injection). The powder is fed by injectors with the help of carrier gas, which
is usually argon.
Figure 1. Schematic of atmospheric plasma spray process (Sulzer Metco) [30]
The powder particles are heated up by the plasma and accelerated towards the substrate to be
coated. The particle velocities typically vary between 100 and 500 m/s depending on the
process. The particles interacting in flight with the plasma jet achieve a molten or semi-molten
state and have velocities sufficient to enable spreading on the interface of the substrate or
previously deposited coating layer. As a result of continuous motion of the spray gun over the
substrate, typically following a meander-like pattern, a uniform coating thickness and structure
can be achieved. The properties of the final coating layer such as thickness, porosity, and
thermal and mechanical properties are controlled by deposition process parameter.
5
3.2. Operating parameters in thermal spraying
The spray parameters and substrate conditions determine the microstructure, physical properties
and adhesion of the coating. On the other hand, the amount and distribution of the coating
material in the spray pattern is determined by the distribution and in-flight characteristics of the
powder in the plasma jet.




Coating material.
Injections of materials.
Energy source.
Kinematic parameters.
Figure 2. operating parameters in thermal spray
Thermal sprayed coatings are built path by path and layer by layer according to the continuous
motion of the spray gun over the substrate surface. The coating properties depend on the
particular process settings, which will be discussed in detail for APS process in this chapter.
Furthermore, the kinematic and geometry parameters related to the spray gun motion
substantially affect characteristics of the final coating layer on a particular component. The
kinematic parameters, such as surface speed of the gun, spray distance and angle relative to the
substrate surface, influence deposition temperature and mass distribution within a coating layer.
These parameters are controlled by the motion of the spray gun typically attached and driven
by the industrial robot (IR):
1.
2.
3.
4.
Relative surface speed between the torch and substrate.
Projection distance.
Projection angle.
Path offset.
6
The quality of the deposit and the temperature of the substrate depends extremely in the
kinematic parameters cited above.
3.3. Relative velocity torch/substrate
The relative torch-substrate velocity expresses the velocity of the point of impact of the jet axis
of molten particles from the torch with the surface of the substrate [16]. It is one of the most
important parameters that strongly influences the characteristics of the deposit on the substrate.
During the thermal spraying process, the molten or semi-molten powders are deposited on the
substrate surface with the torch moving. According to the work of Trifa and al. [11], [16] and
Döring et al. [12], the relative velocity does not change the overall shape of the deposit, but
rather the amount deposited and the maximum thickness. Consequently, per unit time, the
lamella is thicker when the speed of the torch is lower, but at the same time it involves a local
increase in temperature inducing higher residual stresses.
The relative torch-substrate velocity value can be specified by the robot program. In most cases,
it is defined as a constant parameter in the thermal spraying process. Typically, a relative torchsubstrate velocity value of about 340mm/s to 510mm/s is employed [13]. However, when the
distances between two neighbouring scans are not identical, the choice of using a constant
velocity will imply a phenomenon of undulation on the deposit thickness which will be visible
on its surface.
If the motion of the robot still accurate on the trajectory geometry that does not mean necessarily
that the kinetic and dynamic parameters of the tool will follow linearly the trajectory especially
when we meet huge acceleration. So optimizing the root path of the robot is an important feature
to achieve a well coating surface by keeping the same velocity all over the substrate and the
path offset should be adapted to the shape of the workpart we will discuss below the case of
rotated part.
Figure 3. Thermal spray on rotated part
7
The displacement velocity of the torch coupled to the rotation of the part describe the amount
of the deposit on the part in addition the linear velocity of the torch relative to the local radius
and angular velocity of the part would give the above relation
𝑉𝐿 = 2𝑝𝑅∅𝑉𝐴 .
Where 𝑉𝐿 define the linear speed of the torch [mm𝑠 −1 ] and 𝑉𝐴 is the angular velocity [rot/min].
We can observe that the velocity of the torch should be adapted to the value of the radius. We
can conclude that the path offset rather decrease when the velocity of the projection increase.
So we assume that the torch velocity depend linearly with R° we shall rewrite the relation in
this way:
𝑅0
𝑉𝑡 =
∅
𝑅
It is important to maintain and control the projection velocity during the process
3.4. Spray distance
The projection distance is the distance between the outlet of the nozzle, ie. the last geometric
plane of the torch, and the tangent external to the substrate at the point of intersection with the
geometric axis of the torch [16]. The projection distance is one of the most influential
parameters because it affects the temperature and velocity of the particles at impact as well as
the heat flux transmitted to the substrate [25]. The expansion of the gas flow makes it possible
to reach temperatures and maximum speeds (between 800 and 1200 m/s) at the outlet of the
nozzle.
Figure 4. Downstream distance
8
These two parameters gradually decrease as they move away from the nozzle outlet of the torch.
If the substrate is too close, the temperature will be too high at the surface and the residual
stresses will be very strong in the deposit which will damage it. If it is too far away, the particles
will have a reduced speed and, moreover, they will have cooled down and will therefore have
difficulty spreading out properly, as there is a risk of bouncing off the surface. Indeed, during
their flight time, the convective/radiative transfers strongly influence their state, which thus
influences the yield of the deposit and its adhesion to the substrate [26], [27].This is shown in
Figure 1.6. In plasma projection, the projection distance can for example vary from 115 mm to
135 mm [16]. Depending on the materials and their granulometry, the distance is adjusted in
order to improve the deposition efficiency.
3.5. Projection angle
The projection angle represents the angle between the axis of the torch and the tangent at the
point of impact on the surface to be coated. The objective of studying this parameter is to
analyze the shape of the singular strand corresponding to this angle of projection and then to
establish its relationship with the deposit yield. When the angle of projection is 90 degrees, the
spraying takes place like a cone and the deposition on the flat surface normal to the axis of the
jet thus forms a circle whose radius constitutes the width of pass [14], [15].
Figure 5. Thickness distribution with diffrent angle values
presents the evolution of the experimental profiles of the cords as a function of the projection
angle for certain parametric configurations [16]. According to the conclusion of Trifa et al. [16],
9
when the projection angle is closer to 90 degrees, the area of the singular strand of the deposit
is larger. Shows that the thickness of the deposit decreases with the decrease in the projection
angle. Consequently, the yield of the deposit decreases in particular when the angle of
projection is less than 60 degrees. This phenomenon is also demonstrated experimentally in
other works [17], [18].
Normally, a projection angle of 90 degrees is recommended during the thermal spraying
process. The diameter of the largest particles is significantly influenced by the projection angle.
The thickness of the deposit remains constant when the projection angle changes from 90 to 60
degrees and then decreases rapidly when the projection angle decreases to 30 degrees [19], [20],
[21]. In order to obtain a thickness of the homogeneous deposit, the projection angle is generally
maintained as a constant value during the thermal spraying process. Also, the adhesion of the
coating to the substrate is a function of the projection angle: its maximum is obtained between
60 and 90 degrees.
However, when the surface to be coated is too complex or a possible physical collision can
occur between the torch and the workpiece, it is obligatory to readjust the projection angle to
suit the movement of the robot. [22], [23], [24].
3.6. Offset path
The trajectory of the spray gun at the substrate surface represents a sequence of parallel or semiparallel lines. The resulting thickness of the coating layer is accumulated by summation of the
thicknesses of each individual profile. An accurate value should be set to perform a smooth
coating surface without overheating the substrate in case of low offset distance between the tool
passage. Practically in APS applications the value of the offset vary between 5mm and 15mm.
Figure 6. Wavering of the deposit along the path offset
For thermal spraying applications, uniformity of coating can be achieved by specifying a
constant relative torch-substrate velocity and a scanning step equal to or less than a standard
deviation of the singular bead (based on the symmetrical distribution of its assimilated shape
To a Gaussian). In the case where the scanning step is greater than a certain standard deviation,
the undulation on the coating surface is very obvious.
10
4. MODELING, ANALYSIS AND CONTROL OF ROBOT MANIPULATOR
Nowadays the use of robots for the application of the thermal spray coatings is becoming a
standard technique in modern manufacturing industries. Industrial robots enable human
operators to be replaced, and achieve a high level of process accuracy and reproducibility for
diverse processes such as machining [29], arc welding [30] and finally for thermal spray
coating. Furthermore, the use of robots for the coating processes enables operation in vacuum
or hazardous environments of production spray booths. The manipulator model used to perform
the process is The KUKA KR 15/2 available in the laboratory of the university and suits to
achieve the desired goals set by the process.
4.1. System description
KR15/2 robots are six axis industrial robots with articulated kinematics for all continuous path
controlled tasks. The main areas of application are:
 handling,
 assembly,
 application of adhesives, sealants and preservatives,
 machining,
 MIG/MAG welding,
 thermal spray,

YAG laser beam welding.
Figure 7. Main links and joints of the robot
11
All the main bodies are made of cast light alloy. This design concept has been optimized by
means of CAD and FEM with regard to cost-effective lightweight construction and high
torsional and flexural rigidity. As a result, the robot has a high natural frequency and is thus
characterized by good dynamic performance with high resistance to vibration. All the axes are
powered by brushless AC servomotors of plug-in design, which require no maintenance and
offer reliable protection against overload. The main axes are lifetime lubricated, i.e. an oil
change is necessary after 20,000 operating hours at the earliest. The robot can also be quickly
replaced as a complete unit without any major program corrections being required. Overhead
motion is possible. These and numerous other design details make the robots fast, reliable and
easy to maintain with minimal maintenance requirements.
Occupy very small floor space and can be located very near to the workpiece on account of the
special structural geometry [31].
4.2. Robot mechanics
Robot Manipulators are composed of links connected by joints to form a kinematic chain. Joints
are typically rotary (revolute) or linear (prismatic). A revolute joint is like a hinge and allows
relative rotation between two links. A prismatic joint allows a linear relative motion between
two links. For example, a three-link arm with three revolute joints is an RRR arm. Each joint
represents the interconnection between two links. We denote the axis of rotation of a revolute
joint, or the axis along which a prismatic joint translates by 𝑧𝑖 if the joint is the interconnection
of links i and i+1. The joint variables, denoted by θ for a revolute joint and d for the prismatic
joint, represent the relative displacement between adjacent links [32]. In our case study the
KUKA KR15/2 is an articulated RRR model each axis is driven by a transistor controlled, low
inertia AC servomotor. The brake and resolver are space efficiently integrated into the motor
unit. The common arrangement of joints axis z2 is parallel to z1 and both z1 and z2 are
perpendicular to z0. This kind of manipulator is known as an elbow manipulator. The revolute
manipulator provides for relatively large freedom of movement in a compact space. The
parallelogram linkage, although typically less dexterous than the elbow manipulator,
nevertheless has several advantages that make it an attractive and popular design. The most
notable feature of the parallelogram linkage manipulator is that the actuator for joint 3 is located
on link 1. Since the weight of the motor is born by link 1, links 2 and 3 can be made more
lightweight and the motors themselves can be less powerful. Also the dynamics of the
parallelogram manipulator are simpler than those of the elbow manipulator, thus making it
easier to control.
4.3. The workspace
The workspace of a manipulator is the total volume swept out by the end eff ector as the
manipulator executes all possible motions. The workspace is constrained by the geometry of
12
the manipulator as well as mechanical constraints on the joints. For example, a revolute joint
may be limited to less than a full 360° of motion.
Figure 8. Swept volume of the KUKA KR 15/2
The workspace is often broken down into a reachable workspace and a dexterous workspace.
The reachable workspace is the entire set of points reachable by the manipulator, whereas the
dexterous workspace consists of those points that the manipulator can reach with an arbitrary
orientation of the end-eff ector. Obviously the dexterous workspace is a subset of the reachable
workspace.
13
4.4. Axis data
The axis data are visible in the table below show in all the possible motions performed by joints
[33].
Table 1. KUKA KR 15/2 robot joint parameters
Axis
Range of motion
Speed
1
±185˚
152 ˚/s
2
+115˚ to 55˚
152 ˚/s
3
+70˚ to 210˚
152 ˚/s
4
±350˚
284 ˚/s
5
±135˚
293 ˚/s
6
±350˚
604 ˚/s
4.5. Main programing methods
The simulation of the industrial robots before their installation is essential, and to be done, it is
compulsory to go through several stages, geometric modeling shall be the basis to develop a
mathematical approach, follows the modeling under a CAD interface, Analysis of the results of
the simulation. In this chapter we will talk about modeling, CAD, and robot programming in a
general way to explain the mathematical and computational tools that will be used in our study,
and the essential points that we will discuss are:
1. Geometric modeling
 forward and inverse kinematic.
2. CAD and robotics
 simulation.
3. Robot programming
 online
 offline.
4.6. Modeling
The design and control of robots requires the calculation of certain mathematical models, such
as transformation models between the operational space (in which the situation of the end
effector is defined) and l (In which the configuration of the robot is defined).
We distinguish the direct and inverse geometric models which express the situation of the end
effector according to the articular variables of the mechanism and vice versa. The direct and
14
inverse kinematic models which express the speed of the effector as a function of the articular
velocities.
The dynamic models defining the equations of motion of the robot, which make it possible to
establish the relations between the pairs or forces exerted by the actuators and the positions,
speeds and accelerations of the joints
The mathematical formalism uses homogeneous transformation matrices of dimension (4x4).
The homogeneous matrix 𝑗𝑖𝑇 represents the transformation allowing to pass from the reference
𝑅𝑗 to the reference 𝑅𝑖 :
𝑖
𝑖
𝑖
𝑖
𝑖
𝑖
𝑗𝐴
𝑗 𝑃 ] = [ 𝑗 𝑠 𝑗 𝑛 𝑗 𝑎 𝑗 𝑝]
=[
0 0 0 1
0 0 0 1
𝑖 𝑖 𝑖 𝑖
Where 𝑗𝑠 𝑗𝑛 𝑗𝑎 𝑗𝑝 denote respectively the unit vectors along the axes 𝑥𝑗 , 𝑦𝑖 and 𝑧𝑗 of the
𝑖
𝑗𝑇
reference frame 𝑅𝑗 expressed in the reference frame𝑅𝑗 , and where 𝑗𝑖𝑃 is the vector expressing
the origin of the reference frame 𝑅𝑗 in the reference frame 𝑅𝑖 . The vectors 𝑗𝑖𝑠 𝑗𝑖𝑛 𝑗𝑖𝑎 𝑗𝑖𝑝 of the
orientation matrix 𝑗𝑖𝐴 are the directing cosines
4.7. Geometric description
A simple open structure is composed of n + 1 bodies denoted 𝐶0 , ..., 𝐶𝑛 and n joints. The body
C0 designates the base of the robot and the body Cn the body which carries the terminal organ.
The joint j connects the body 𝐶𝑗 to the body𝐶𝑗−1 . The method of describing DENAVITHARTENBERG (DH) is based on the following rules and conventions: • bodies are supposed
to be perfectly rigid. They are connected by articulations considered as ideal (no mechanical
play, no elasticity), either rotoids or prismatic; The reference frame 𝑅𝑗 is linked to the body 𝐶𝐽 ;
 the axis 𝑍𝑗 is carried by the axis of the articulation j;
 the axis 𝑥𝑗 is carried by the perpendicular common to the axes 𝑍𝑗 and𝑍𝑗+1.
If the axes zj and 𝑍𝑗+1 are parallel or collinear, the choice of 𝑥𝑗 is not unique: considerations of
symmetry or simplicity then allow a rational choice. The transition from the reference 𝑅𝑗−1 to
the reference 𝑅𝑗 is expressed as a function of the following four geometric parameters:
 𝛼𝑗 : angle between the axes 𝑍𝐽−1 and 𝑍𝑗 corresponding to a rotation around𝑥𝑗−1 ,
 𝑑𝑗 distance between 𝑍𝐽−1 zj-1 and 𝑍𝑗 along 𝑥𝑗−1 ,
 𝜃𝑗 : angle between the axes 𝑥𝑗−1 and 𝑥𝑗 corresponding to a rotation around𝑍𝑗 ,
 𝑟𝑖 : distance between 𝑥𝑗−1 and 𝑥𝑗−1 along 𝑍𝑗 .
The transformation matrix is defined in the following expression:
𝑖
𝑗 𝑇 = 𝑅𝑜𝑡(𝑥, 𝛼𝑗 )𝑇𝑟𝑎𝑛𝑠(𝑥, 𝑑𝑗 )𝑅𝑜𝑡(𝑧, 𝜃𝑗 )𝑇𝑟𝑎𝑛𝑠(𝑧, 𝑟𝑖 )
15
Where Rot (u, α) and Trans(u, d) are homogeneous transformation matrices (4x4) respectively
representing a rotation α about the axis u and a translation d along u.
The forward model geometric model is the set of relations which allow to express the situation
of the end effector, that is to say the coordinates Operation of the robot, as a function of its
articular coordinates. In the case of a simple open chain, it can be represented by the
transformation matrix 𝑛0𝑇:
𝑖
0
𝑛−1
𝑛𝑇 (𝑞𝑛)
𝑗 𝑇 = 1𝑇(𝑞1) × … ×
The inverse geometric model (MGI) allows to express the articular variables q of the
manipulator arm according to the operational coordinates X required for the execution of a
given task. It is the set of relations inverse to those of the direct geometric model.
𝑞 = 𝐹 −1 (𝑋)
There is no general analytical method to solve the inverse kinematic model. However, a number
of methods, adapted to particular classes of kinematics, are often cited in bibliography and allow
to deal with the problem.
The Pieper method: it is adapted to manipulating arms with six degrees of freedom with three
rotating articulations of concurrent axes or three prismatic joints. It is widely used because
virtually all the manipulator arms have a wrist with 3 concurrent axes.
Paul's method: it treats each particular case separately and is suitable for most industrial
manipulator arms. It is a heuristic method that does not admit a deterministic procedure.
The method of Raghavan and Roth gives the solution for a 6R robot from a characteristic
polynomial of degree 16 (16 solutions).
Numerical methods: The transformation of coordinates can be obtained by numerical
procedures, by successive iterations (resolution of a system of nonlinear equations).
Figure 9. Possible solutions for a single position
16
The number of possible solutions of the inverse geometric model depends on the desired
situation, the number of degrees and the morphology of the manipulator arm. There are
practically three cases:
1. No solution when the desired situation is outside the working space of the manipulator
arm.
2. Solution in finite number when all solutions can be calculated unambiguously (8
solutions in the case of manipulator arms with six degrees of freedom having six rotary
links, three of which have concurrent axes).
3. Infinity of solutions.
The kinematic scheme is the tool that gives us the possibility to simplify and schematize the
relative movements between the bodies of the robot, in order to study its functioning.
The architecture of KR 15/2 robot is composed of 6 bodies and 6 joints. By associating to each
body Ci a reference Ri with an axis Zi borne by the axis of the articulation Ai, the following
kinematic scheme is obtained:
Table 2. Denavit-Hartenberg parameters of KUKA KR 15/2
Joints
𝜶
𝒅
𝝑
𝒓
𝑱𝟏
0
0
𝑞1
0
𝑱𝟐
−
𝐿1
𝑞2
0
𝑱𝟑
0
𝐿2
𝑞3 −
𝑱𝟒
−
0
𝑞4
𝐿3
0
𝑞5
0
0
𝑞6
0
𝑱𝟓
𝑱𝟔
𝜋
2
𝜋
2
𝜋
2
𝜋
−
2
𝜋
2
0
L1= 300mm , L2= 650mm , L3= 600mm.
The elementary transformation matrices
By replacing the parameters DH in the global form of the matrix of passage between the marks,
one obtains the matrices of the following elementary transformations:
17
The above results show the orientation and the position of the wrist expressed in the base frame
of KUKA robot used in this work and that’s the basis used by most of the robotic software to
generate motion by inverse kinematic not more details of computation are figuring in this thesis
since RobCAD compute all the motions parameters automatically the critical part of
computation will be done mainly for the orientation and position of the torch during the process.
4.8. Robotic CAD systems
Robotic CAD systems offer powerful graphical tools to deal with problems such as:

Choice of the most suitable robot for a given task.

Place robots in a site.

Avoidance of collisions between the robot and the working environment.

Generation of optimal trajectories.
18

Calculation of the cycle time and its minimization.

Optimize the process
The management of a robotic site must pass through a Robotic CAD system whose goal is to
solve all problems related to their operation. The elements encountered to do a graphical
modeling and simulation are the robot, the task and the working environment. Robotic CAD
systems are generally used for three purposes:
1. The modeling and construction of the robot and its environment, starting from the
creation of elementary geometric shapes, first, and subsequently complex shapes By
transformations such as the intersection, union, substitution, etc., of existing forms.
2. The generation of the movement and the programming, which allows to program the
tasks that will be carried out by the robot.
3. Simulation which allows to animate the robotic site and to detect the possible risks of
the collisions. The simulator must read all the data necessary and representative of the
reality (relativity robot / task, robot / work environment ... etc).
PLM software systems are mainly used for the following characteristics:
 Modeling of the site.
 Representation closer to reality.
 CAD-site coupling, with the possibility of information retrieval or not.
 Ease of development: user interface.
 Interaction with other CAD systems.
 Analysis and optimization tools.
 Collision detection.
 Modification of the robot environment, if necessary.
 Generation of motion without collision.
4.9. Control of the robot
Robots and like any other automatic system must be programmed, and their programming can
be done in several ways, by manual teaching, offline or auto-teach. Dedicated applications
software (e.g. painting, welding, etc.), which takes into account the specificities of the process,
are available for different applications.
19
4.9.1. Program by teaching on-line
The on-line programming of the robots is typically carried out by a skilled operator. The
operator guides the robot through a desired path, which is usually marked directly on a surface
of a set-up component. For example, for a simple geometry, the spray gun trajectory represents
a meander-like pattern, created by application of parallel paths with a constant distance between
them and with a constant gun speed. The robot motion in the teach-in method is controlled
manually by a teach pendant which causes the naming of this technique. The programming
involves manually jogging the robot between selected locations which are placed on the marked
path. These locations have to be recorded in the robot controller memory to be later utilized for
a continuous motion of the robot.
Figure 10. Online programming
4.9.2. Offline programming
The main specific feature of the off-line programming (OLP) methods is the performing of
program development operations not in the production booth but in a test environment with
subsequent transfer of the program to the robot. The off-line environment can be represented
by another test booth used only for the programming efforts, or by a virtual copy of the
production booth created with simulation software. The hardware-based off-line methods avoid
blockage of the production booth but have the same technical drawbacks as conventional online methods discussed above, caused by using on the same teach-in approach to generate the
robot program. These methods are widely used in the automation industry for robotic systems
and overall production lines for example for car parts.
20
Figure 11. Off-line programming in RobCAD
There are various software tools such as RobCAD from Siemens PLM Software, RobotStudio
from ABB, Delimia from Dassault Systems etc., IGRIP from Deneb Robotics, etc. which were
developed for off-line programming
The advantages of this approach are listed below:
 To simulate graphically the robot program (view of the robot in motion on the screen,
checking the cycle time).
 Create or edit programs.
 Transfer robot programs from the computer to the control cabinet.
There are two main approaches for offline robot programming: robotic programming and
graphical programming.
4.9.3. Programming by robotic language
Robot programming languages are very similar to traditional informatic languages. They are
highly efficient in defining positions by calculations or to implement conditional statements or
loops. The main disadvantage of these languages is that they generally remain specific to each
brand of robot, and there is no recognized standard. The following languages can be cited:
RAPID (ABB), VAL II (UNIMATION), RAIL (AUTOMATIX).
4.9.4. Graphical programming
This method and more convenient to program robot trajectories on complex parts. Current
CAD/CAM software is used to represent and model robots, parts, and the task environment.
21
This offline programming method offers undeniable advantages over the language approach
such as the visualization of work in progress thanks to the interfaces of the CAD / CAM
software, and the possibility of creating geometric databases for robotic cells
4.9.5. Programming by self-learning
A new patented function of self-learning path by the robot for finishing operations such as
polishing, deburring, sanding, grinding ... also exists. This solution, which combines robot
effort control and user interface simplifies, significantly reduces the time and cost associated
with programming, and can be used for all contact processes (metallurgy, mechanics, foundry,
plastics…).
22
5. EXPERIMENTAL EVALUATION OF PROCESS PARAMETERS AND MODEL
VERIFICATION
This chapter presents a mathematical model to estimate the thickness of the final deposit with
respect to the kinematic parameters of the robot. It also makes it possible to simulate the profile
of the deposit, and to calculate the flatness of the deposit surface. The effects of the kinematic
parameters on the thickness of the deposit and the homogeneity of the deposit will be analyzed,
it is therefore possible to obtain a preconceived deposit. The increasing demands on coating
productivity lead to the analysis of the effects of robotic kinematic parameters.
We shall detail later the experimental protocol in order to plan the step of the modelisation of
the process before implementation in the virtual robotic cell
5.1. Characterization of Deposit Thickness
The deposition is the result of the superposition of a multitude of random and independent
events: the impacts of the incident molten particles on the substrate [4]. Even if the impacts of
the particles are random and independent events, their result is a deposit with a defined shape
that can be described by a continuous probability function [5], [6]. The process of thermal
projection can be considered as the accumulation of singular strings (a projection pass) of
deposition.
The strategy of this simulation is based on the model of the final deposition which is based on
the mathematical function described not only by the singular cord model but also by the
kinematic parameters [7], [8]. The kinematic parameters considered in this chapter are: the
projection distance, the number of passes and the scanning step. The combined effects that are
caused by their interactions together determine the properties of the coating (eg, the thickness
of the deposit). It is therefore necessary to optimize the trajectory of the robot for the thermal
spraying process in order to obtain the thickness of the desired deposit.
5.2. Modeling of the singular strand of the deposit
The profile of the strand is greatly influenced by the property of the material, the projection
distance, the projection angle and the deformation of the substrate caused by local heat transfer.
Numerous publications have shown that the coating profile can be characterized by a
mathematical formula [9], [10], [11], which offers the possibility of analysis and optimization
of the effect of the kinematic parameters by this mathematical method. The analysis of this
study is based on the Gaussian distribution
23
5.3. Mathematical model of the singular spray
The definition of the mathematical model of the singular cord requires independent variables
that define the profile of this model. They contain: the maximum height of the singular cord, its
width and its area in 2D The Gaussian function which represents the profile of the singular bead
is expressed by:
𝑍(𝑥) =
𝐾
𝜎√2𝜋
𝑒𝑥𝑝 [−
(𝑥 − 𝜇)2
]
2𝜎 2
Where Z (x) represents the probability density of the normal law but also the height of the
singular bead at the abscissa position x. Σ is the standard deviation of the Gaussian distribution,
it also quantifies the width of the singular strand. Μ is the mean of the heights. K is a constant
coefficient.
Figure 12. Mathematical model of the footprint of the spray
5.4. Mathematical description of the singular spray profile
The singular strands are formed by the cold spraying system for different spacing distances: 10
mm, 30 mm, 50 mm and 70 mm. It is therefore necessary to quantify the mathematical
relationship between the profile of the singular cord and the distance of projection.
The profiles of the singular spray are measured five times. The height of the singular strand is
calculated as the mean value of these five measured values. A set of points taken from different
strands at different projection distances is introduced into MATLAB. It is thus represented by
five Gaussian curves, illustrated by the above figure and shows that the change in maximum
heights at different projection distances is significant. There is an obvious increase in the height
of the singular spray when the distance of projection increases from 10 to 50 mm and a marked
decrease in the height of the singular cord when it increases from 50 to 70 mm. In this
experiment, the maximum height of the singular cord appears at a projection distance of 50 mm,
24
at a height of 1.325 mm. The minimum height of the strand is encountered at a projection
distance of 10 mm, at a height of 0.8 mm [29]. There is therefore a two-degree relationship
between the projection distance and the maximum height of the singular spray.
5.5. Experimental protocols
Figure 13. Procedures for generating an offline trajectory [30]
25
5.6. Acquisition of the CAD file
In this work CAD models were designed by NX CAD system in order to complete the missing
parts of the workcell in RobCAD. CAD translator was used to import parts from NX to the
workcell layout to be part as functional elements of the process such a the torch and its support
mounted on the writ of the robot the tool holder to fix the parts on the rotary table and a flat
surface to test the correctness of the spraying process.
 First the active tool of the spraying process was design with three different parts as the
support mounting flange on the wrist of the robot then a second support to hold the torch
during application and keep it stable during the motion of the robot in order to guarantee
an accurate projection on the substrate then assembled under dimensional condition of
the robot wrist that was found in the robot data sheet:
Figure 14. Mounting flange of the KUKA KR 15/2 [34]
Figure 15. Mounting data of the KUKA robot [34]
26
The above drafting information of the KUKA robot are the basic of design and the mounting of
the torch on the robot wrist, the first support was designed to fit the robot mounting condition.
Figure 16. Support mounted on the robot wrist

Then a second support was designed to hold the torch from the nozzle that was built by
assembly constraints in order to erase all degree of freedom and fix the relative motion
between the parts and the wrist of the robot. Hereby you can see the final assembly of
the parts in NX. The torch used in this study is an F4 plasma spray gun this machinemounted, multi-purpose air plasma spray gun used for external coating applications.
The gun is capable of operating at power levels up to 55 kW equipped with the standard
nozzle of 6 mm, virtually all spray parameters can be met with all the equipment
mounted the weighs of the torch is approximately 9.7 Kg. After modeling the TCP
should be set to the robot flange coordinate system which is not at the centre point of
the wrist. Therefor we shall use homogenous transformation matric to position the tool
with the robot.
Figure 17. Final assembly of the thermal spray gun
27

The next generation of CAD data is composed of downloaded parts that are attached by
a tool holder designed to fit the lower part of the turbine blade then assembled in NX
by constrains to their functional position. After investigation of the overall dimension
of the turbine blade the Tool holder was designed and placed on the work table by the
assembly constraint in NX through the central hole of the work table to deny all degrees
of freedom during the process.
Figure 18. Turbine blade mounted on the rotary table
The studied part is a rotary turbine blade that represent a structural material and component of
gas turbine engines for aircraft and power generation applications. Operate under very
aggressive conditions characterized by high temperature and mechanical load in an oxidizing
and corrosive atmosphere. The inlet temperatures in stationary gas turbines are about 14001500°C [30] and tend to exceed 1600°C for very modern engines. Increasing demands for
turbine performance and efficiency require an increase in inlet temperatures up to levels close
to melting points of the typical structural materials. Rotating blades and stationary vanes are
the most loaded parts of the turbine, which are working at very high temperature and thermomechanical stresses. Furthermore, the rotating blades are subject to extreme centrifugal forces,
which they have to sustain at high temperature in an aggressive environment. The combustion
chamber and hot section of the turbine consist of diverse components which provide structural
integrity and at the same time the thermal and environmental protection of the turbine
components.
28
Figure 19. Overall dimension of the Turbine blade

For the parts (substrates) to be coat the generation the meshing of the geometry is a
critical step to simulate; in FEM simulation module available in NX allowed us in this
case study to generate a surface meshing of the workpiece. The first step is to generate
the model then import it to the FEM platform and generate a surface 2D meshing with
triangle shell 281 element comprising 4 nodes. It is well adapted to model the shell
structure which is linear and deformed [30]. This element having six degrees of freedom
at each node, its geometrical deformation is linear in both directions within the
framework and represent a smooth creation of robot trajectories point by point on the
substrate surface.
Figure 20. Meshing the workpiece under NX Nastran
29
5.7. Selection of the thermal spray operating parameters
Thermal spray operating parameters are classified into several groups according to published
studies: the energy parameters, powder injection parameters and various kinematic parameters.
These parameters can be directly controlled (speed of the torch, projection distance, scanning
step, etc.), or indirectly controlled (speed and temperature of particles in flight, etc.). They can
influence the performance of projection and the final coating properties. Many publications [28]
described the relationship between operating parameters and coating characteristics as well as
coating structures. A Kout et al. [29] investigated the planning path-oriented spray-coating
processes, they represented an optimization method to compute and approximate the desired
coating thickness with coating relative parameters. M.M. Fasching et al. [29] presented an
approach for spraying layers using robotic thermal spraying system, they offered equations to
optimize the spray angle, to generate more accurate robot trajectory. F. Trifa et al. [30] studied
the interaction between the operating parameters and characteristics of the deposit, which
allows to select the proper settings. S. Guessasma et al. [29] have developed an intelligent
system based on fuzzy logic to assist the choice of parameters depending on the desired
characteristics and desired deposition. Therefore, operating parameters should be carefully
chosen and kept constant during thermal spray process in order to lead to desired and optimized
coating properties.
Table 3. Example of the kinetic parameters of turbine blade coating
A=45%
Nominal deposition efficiency:
M =50 g/min
Powder feed rate:
Spraying time:
∆t =2 sec
Nominal spray distance:
150 mm
Nominal standard deviations :
𝜎x =4 mm, 𝜎y =3 mm
Nominal spot displacements :
∆𝑥0 =4 mm, ∆𝑦0 =2 mm
Nominal angle of spot rotation :
𝜎0 =30°
Angle of jet divergence :
𝛾 =15°
𝜌𝑐𝑜𝑎𝑡 =5,5 g/cm3
Coating density: coat :
Spray angle efficiency factor :
m=0,5
Distance efficiency factor :
q =0,5
30
5.8. Trajectory generating
This method consist of generating the mesh of the CAD model in this case study This study
proposes the use of a SHELL281 element with 4 nodes generated with the FEM module of NX
Nastran. It is well adapted to model the shell structure which is linear and deformed [4]. This
element having six degrees of freedom at each node, its geometrical deformation is linear in
both directions within the framework of the plane, then find out the trajectory from the mesh
points and normal vectors to the surface. There are many file formats of geometric models
available for this operation, such as STL, IGES, STEP, ASCII, and ACIS etc. The STL file
describes the workpiece with triangular mesh and every triangle is defined by three vertices and
a normal vector to the surface in three-dimensional Cartesian coordinate system. To select the
points on scanning curve and calculate the torch orientation, the vertices and normal vectors are
required. The STL format is the most appropriate for this method.
In this work CAD models were developed by NX CAD system from Siemens in order generate
the meshing of the geometry with the FEM simulation module available within the CAD
software of the workpiece the first step is to generate the model then import it to the FEM
platform and generate a surface 2D meshing with triangle shell 281 element comprising 4
nodes. It is well adapted to model the shell structure which is linear and deformed. This element
having six degrees of freedom at each node, its geometrical deformation is linear in both
directions within the framework and represent a smooth creation of robot trajectories point by
point on the substrate surface.
After modeling the part we shall import the data through a converter to IGES file into the virtual
workcell of the robot with the adequate format. Then start the cell construction by using the
placement command to put the elements in their process places before starting the generation
of the trajectory of the projection within RobCAD.
31
6. INTRODUCTION TO OFFLINE PROGRAMMING AND COATING SIMULATION WITH
ROBCAD SOFTWARE
Implementation of the RobCAD software enables to perform robot path planning and coating
thickness simulation within the virtual cell. The virtual cell in RobCAD represents a CAD
model of the actual spray booth. The virtual model of the booth with equipment for APS coating
processes including robot manipulator with additional rotational axes, spray guns and
components to be coated with corresponding tooling and fixtures is established and introduced
into the graphical simulation system of RobCAD.
6.1. Spray robot programming in RobCAD
The RobCAD software package includes the Paint module, which was originally developed for
simulation of painting process. The coating thickness distribution can be simulated by the Paint
module with an input of the thickness distribution in a basic paint profile. The ability of the
software to simulate an accumulation of thickness, provided by application of an arbitrary
number of spray profiles, allowed RobCAD to be applied with the Paint module for the
simulation of thermal spray process. In order to enable a robotic simulation in combination with
coating simulation, the RobCAD software should be set up and correctly configured. Following
configuration steps must be performed to set up the work cell.
6.1.1. Workcell layout
Set up of the KUKA robot model in RobCAD work cell and attachment of the spray gun model.
In addition auxiliary manipulators such turntables must be imported and programmed as
external robotic axes to perform motion of the component holder. The part has been fixed with
NX Assembly module by using constraints the two below parts was downloaded.
Figure 21. The robot worcell
32
The spray gun is mounted on the robot flange at the 6th axis by using the placement command
that ensure an accurate placement and position of the tool in this work a special frame situated
on the TCP of the torch was created in order to inform the robot of the tool position and
orientation then using the attach constraint to keep the same distance and orientation of the tool.
Figure 22. The mounting the spray gun on the robot flange
The additional frames in the above figure were set in the motion command in RobCAD the aim
of it is to transform the wrist centre point which carry the end effector frame to the output of
the torch with the Z axis pointed out in order to configure it during the spray process to control
the mass distribution of the particle on the substrate surface then finally evaluate the thickness
of the coating part in the paint module of RobCAD then the TCP of the spray gun is set in
mounting options as an external axis of the robot.
The technical data of the process are also imported from the FEM platform of NX in order the
visualize the meshing of the coated surface a critical data of generating the trajectory a CAD
translator is also needed to transfer the results to the worcell in RobCAD then placed to their
functional position with the placement editor. Input of the CAD model of the component into
the RobCAD environment. Typical format of the model is an IGES file.
33
Figure 23. The mesh generation of the turbine blade
Completing the exact CAD models of the component holder and needed auxiliary tools. The
modeling could be done with the internal RobCAD design module or with the help of the
external software with a subsequent import into RobCAD. The holder, tools and a component
must be assembled together and attached to the turntable.
Figure 24. The fixation of the working part
34
6.1.2. Path editor
path editor is a tool available in RobCAD that enables us to create locations and orientations on
a given control points by using whetear the pick points option or creating location and
orientation manually by setting the value of the position and orientation.
Figure 25. The path editor module
Then create a pathway that is basically the secession of the locations in order to create a
continuous path on the substrate surface. That will appears as a dashed lines oriented by arrows
which shows the motion flow of the path it could also set with linear or circular interpolation
Figure 26. Create path dialog window
35
After choosing the right order of the locations RobCAD generate by using inverse kinematics
to determine the joint values needed to reach a given target location that consists of sequence
of target locations (TCP position and orientation) with associated attributes.
Figure 27. An exemple of a path created in RobCAD
Once the path ready we can run it with the motions command and simulate the motion of the
robot after setting it as active mechanism and at the same time control the linear and angular
speed of the end effector in the motion setting under the operatory parameters of the spray
process.
6.1.3. Calibration and the adjustment of the virtual RobCAD work cell.
The calibration procedure is compulsory to ensure that the offline programmed robotic path
and programmed robotic locations meet their real positions in the actual workcell. These three
locations should be created in RobCAD and downloaded to the booth. The robot with a distance
measurement tool mounted onto the robot or spray gun should be moved to these locations in
the real booth. If the RobCAD location does not match the real position, the robot with the
measurement tool will be moved manually to the correct position. This new manually adjusted
position will be stored together with the initial location in the robotic controller memory for
each of the three locations. Thus, the three pairs of the programmed and adjusted locations will
be created. This pairs of locations will be uploaded back to the RobCAD [30].
36
6.2. Application of the RobCAD/Paint software for coating thickness simulation
The coating thickness simulation is performed by RobCAD software with the reprogramming
of the Paint module according to the model equations developed. The RobCAD software allows
to simulate coating distribution on the part, resulting from arbitrary motion of the TCP frame.
The TCP frame is connected to the spray gun and moves along the robot path following the
programmed locations at selected speed. The spray programs to apply profiles and complete
coating layers onto the flat substrates were generated with OLP technique in RobCAD.
Examples of the robot programs to apply these coating patterns with subsequent thickness
simulation of APS process the trajectory of the TCP frame is shown by the dashed line.
Figure 28. RobCAD thermal spray application
Corresponding target locations are divided by rectangular frames. The stand-off distance and
gun orientation were programmed to stay constant and equal to their nominal values during the
spraying onto the substrate.
After performing of the robot motion simulation, corresponding result of thickness simulation
appears on the substrate surface in the shape of a color map, which defines thickness distribution
at each substrate point. The color scale at the images represents thickness given in micrometres
[30].
37
6.3. Summary of the simulation of coating deposition for Turbine blade
For this session only the Conceptual diagram of coating process development is done the
simulation of the process still hard to perform without a deeper knowledge and control of the
operating parameters for a such complex structure therefore my research work reached only the
protocol of the process with the perspective of improving it in a near future with a deeper
research in thermal spray processes.
Figure 34 Conceptual diagram of coating process development with application of offline
programming and coating thickness simulation in RobCAD software [30]
38
7. CONCLUSION
The aim of this study is to develop methods to apply thermal spray coating by robot application
on complex sample’s surfaces in the requested conditions. If the application on simple surfaces
is well done since long time ago it still complicated even nowadays to apply those methods on
some complicated samples. Long-time of realisation that cost a lot before or after
implementation. This case study introduce new methods by using CAD environments as
RobCAD that allow a realistic approach of the functional data of the process by using the
painting module that enable us to simulate the coating distribution on a free form surface and
techniques to develop and optimize the parameters to apply a perfect coating surface.
Generation of the trajectory on complex parts geometry
Modification and optimisation of the kinetic parameters of the trajectory of the robot (speed,
orientation, path offset) and the torch orientation at each control point.
Create optimized curves on the trajectory
Experiments shows that the trajectory generated on RobCAD are compact with the real
trajectory of the robot. But a divergence was observed while controlling the dynamic response
of the system the projection velocity still uncontrolled over all the surface since some
parameters are not taken into account like the cables and torch weight that induce us to some
divergence between the virtual and real booth. Without controlling the dynamic of the robot the
quality of work is hard to achieve in order to obtain a constant deposit thickness of the materials
and the heat load on the substrate surface. Therefore the dynamic parameters are an important
feature to control and analyse in the real cell in order to achieve a constant projection velocity
on the substrate to ensure an homogenous surface coating.
Real time communication with manipulator in order to verify the position and orientation of the
torch
Generate a continuous trajectory by attaching the control points of the thermal spraying
Compare the real and the virtual trajectory of the robot.
In thermal spraying the capacity to simulate and control the thickness of the coating is an
important parameter in fact the prediction of the thickness enable us to validate or not the
trajectory ( position , orientation , displacement velocity) defined by the Off line programming
if the results are compact with the real thickness distribution the robot programme is efficient.
It was also pointed out by several research project that the working conditions influence
extremely the results.
Several parameters should be fixed and analysed to perform a good combination of the robot
motion and the high velocity projection of the torch during this project work the kinematic
aspect of the process was mainly investigated but still not enough to ensure a good work quality
therefore more deep knowledge on spray technology and heat treatment should be done to
ensure the requested quality of work.
39
REFERENCES
[1]
S. Deng. Programmation robotique hors-ligne et contrôle en temps réel des trajectoires:
développement d’une extension logiciel de RobotStudio™ pour la projection thermique.
Thèse de doctorat. Université de Technologie de Belfort-Montbéliard, Belfort, France,
2006.
[2]
P. Nylén, I. Fransson, A. Wretland. Coationg thickness prediction and robot trajectory
generation of thermal sprayed coatings. Thermal Spray: Practical Solutions for
Engineering Problems, C.C.Berndt (ED), ASM International, Materials Park, Ohio-USA,
1996.
[3]
J. Li. Modélisation de la formation des contraintes résiduelles dans les dépôts élaborés
par projection thermique. Thèse de doctorat. Université de Technologie de BelfortMontbéliard, Belfort, France, 2006.
[4]
Les dernières nouveautés en matière de projection thermique. Technical report,
Conférence internationale et exposition à bâle. Disponible sur: http://www.empa.ch.
[5]
W. Xia, H. Zhang, G. Wang, Y. Yang. A novel integrated temperature investigation
approach of sprayed coatings during APS process. Journal of materials processing
technology 209(2009) 2897-2906.
[6]
Meillot, E.,Bianchi, L.,Roussel, E., Freslon, A.,2001. Industrial applications with the
atmosphere and temperature controlled plasma spraying process. High temperature
materials and processes 5,51-59.
[7]
Moceau, C.,1998. Towards a better control of thermal challenges of the 21st century.
ASM International, Nice, France, pp.1681-1693.
[8]
Moulla, L.,Salhi, Z.,Planche, M.P., Cherigui, M., Coddet, C., Belfort, F.,2005. On the
measurement of substrate temperature during thermal spraying. In : Lugscheider, E.(Ed.),
Thermal spray connects: Explore its Surfacing potential. DVS/IIW/ASM-TSS, Basel,
Switzerland, pp.679-683.
[9]
E. Lugscheider, R. Nickel. Finite element simulation of a coating formation on a turbine
blade during plasma spraying. Surface and coating technology. 174-175(2003) 475-481.
[10] R. Bolot, J. Li, R. Bonnet, C. Mateus, C. Coddet. Modeling of the substrate temperature
evolution during the APS thermal spray process. Thermal spray 2003: Advanceing the
Science&Applying the Technology,(Ed.) C. Moreau and B. Marple. ohio, USA 2003.
40
[11] F. Trifa. G. Montavon, C. Coddet. On the relationships between the geometric processing
parameters of APS and the Al2O3-TiO2 deposit shapes. Surface and Coatings
Technology, Volume 195, Issue 1, pp.54-69, 2005.
[12] J. Döring, R. Vassen, D. Stöver. The influence of spray parameters on particle properties,
Proc. 2002 International Thermal Spray Conference and Exposition, E. Lugscheider, PA.
Kammer (eds), DVS-Verlag GmbH, Düsseldorf, Germany, p. 440-448, 2002.
[13] A. McDonald, P. Eng., K. School, P. Eng. B. Harvey. Thermal spraying training module.
P17, 2008.
[14] P. Chedmail, E. Dombre, P. Wenger. La CAO en robotique. P206-208, 1998.
[15] J. Ilavsky, A. Allen, G.G. Long, S. Krueger. Influence of spray angle on the pore and
crack microstructure of plasma-sprayed deposits. Journal of the American Ceramic
Society, Volume 80, Issue 3, pp 733-742, 1997.
[16] F. Trifa. Modèle de dépôt pour la simulation, la conception et la réalisation de revêtements
élaborés par projection thermique. Thèse de doctorat. Université de Technologie de
Belfort-Montbéliard, Belfort, France, 2004.
[17] M.F. Smith, R. Neiser, R.C. Dykhuizen, in S. Sampath and CC. Berndt (eds.), Thermal
Spray Industrial Applications, ASM International, Materials Park, OH, USA, 1994, pp.
603-608.
[18] G. Montavon, S. Sampath, C.C. Bemdt, H. Herman and C. Coddet, in S. Sampath and
CC. Bemdt (eds.). Thermal Spray Industrial Applications, ASM International. Materials
[19] C. Morceau, M. Lamontagne and P. Cielo, in T.F.Bernecki (ed.), Thermal spray research
and applications, ASM International, Materials Park, OH, USA, 1991, pp.19-26.
[20] G. Montavon, S. Sampath, C.C. Berndt, H. Herman, C. Coddet. Effects of the spray angle
on splat morphology during thermal spraying. Surface and Coatings Technology. 91
(1997) 107-l 15.
[21] S.H. Leigh, C.C. Berndt. Evaluation of off-angle thermal spray. Surface and Coatings
Technology. 89(1991) P213-224.
[22] R. Gadow, A. Candel, M. Floristán. Optimized robot trajectory generation for thermal
spraying operations and high quality coatings on free-form surfaces. Surface & Coatings
Technology. 205 (2010)1074-1079.
[23] D.S. Duvall and D.L. Ruckle. Ceramic thermal barrier coatings for turbine engine
components. ASMEPaper 82-GT-322, p1-9.
41
[24] V.V. Sobolev, J.M. Guilemany. Flattening of droplets and formation of splats in thermal
spraying: a review of recent work- Part 2. Journal of Thermal Spray Technology. Volume
8, Issue 2, pp. 301-316, 1999.
[25] I. Ficher. Variables influencing the characteristics of plasma-sprayed coatings.
International metallurgical reviews. 1972, vol.17, pp.117-129.
[26] 26 C. Li, B. Sun. Effects of spray parameters on the microstructure and property of Al2O3
coating sprayed by a low power plasma torch with a novel hollow cathode. Thin solid
films. 450(2004) 282-289.
[27] K.A. Khor, Y. Gu, D. Pan, P. Cheang. Microstructure and mechanical properties of
plasma sprayed HA/YSZ/Ti-6Al-4V composite coatings. Biomaterials 25(2004) 40094017.
[28] V. Hurevich, A. Gusarov, I. Smurov. Simulation of coating profile under plasma sprayin
conditions. International Thermal Spray Conference 2002 Proceedings. DVS Verlag,
2002, pp.318-323
[29] Programmation robotique en utilisant la methode de maillage et la simulation thermique
du proc´ed´e de la projection thermique Zhenhua Cai pp15- 34 (2011)
[30] Modeling and offline simulation of thermal spray coating process for gas turbine
applications Vom Fachbereich Maschinenbau der Technischen Universität Darmstadt pp
80- 100 2014
[31] Robot Modeling and Control,First Edition,Mark W. Spong, Seth Hutchinson, and M.
Vidyasagar pp 10 26
[32] Jerome Barraquand and Jean-Claude Latombe. Robot motion planning: A distributed
representation approach. International Journal of Robotics Research, 10(6):628–649,
December 1991
[33] S. Deng, H. Liao, C. Zeng, C. Coddet. Robotic trajectory autogeneration in thermal
spraying. Thermal Spray 2005: Explore its potential, (Ed.) E. Lugscheider, ASM.
International, Materials Park, Ohio, USA, 2005.
[34] The KUKA data specification concerning the KR 15/2 model.
42