Report TFM Bertacarbonell

Master Final Thesis
Master’s Degree in Industrial Engineering
Usage of a mechatronics software for efficient machine

development
REPORT
Author: Berta Carbonell i Vilaplana

Director: Pablo Murciego
Academical advisor: Lluïsa Jordi i Nebot
Call: April 2020
Escola Tècnica Superior

d’Enginyeria Industrial de Barcelona
Pg. 2 Report
Usage of a mechatronics software for efficient machine development Pg. 3
Abstract
Industrial machine manufactures face an increasing demand to deliver new high-quality
products in a short time and at a reasonable price. Customers expect to receive machines that
are productive, robust and flexible enough to adapt to their needs. Meeting these
requirements while developing new technologies is a great business challenge. The
development process is complex and with high degrees of uncertainty. Thus, the industry
needs new development methods and tools that help them deal with complexity, be more
efficient, and reduce costs.
These tools are software that emulates physical systems into virtual models. By using
them, engineers can verify machine concepts, integrate designs or validate control
programs to make informed decisions before any hardware is built.
In this context, this project shows a case study where a system is developed using a model-
based approach. The system is a prototype meant to automatically pick 3D printed parts and
place them in a packaging box. The machine model is developed to validate and optimize the
pick and place algorithms before testing them in the physical equipment. The final simulation
can understand the position, speed and acceleration values introduced by the user and provide
a correct 3D visualization of the machine movement.
The model includes the physical properties and kinematics of all the moving machine
components, the behaviour of the motor drives and the data exchange between the controller
and the peripheries. The simulation is governed by the same control program that is executed
in the real machine.
The virtual model is validated by a 3D visualization of the machine motion as well as an

analysis of the motion graphs of several movement sequences. The project also includes an
analysis of the costs and the environmental impact of this work.
This thesis also analyses the advantages and implications of using a virtual model to develop
a machine. This is done through a literature review and observations made during the
development of the case study.
Pg. 4 Report
Table of Contents
TABLE OF CONTENTS _________________________________________ 5
1. GLOSSARY ______________________________________________ 7
2. INTRODUCTION ___________________________________________ 8
2.1. Objectives ....................................................................................................... 8
2.2. Context ........................................................................................................... 9
2.2.1. Usage of programmable logic controllers .......................................................... 9
2.2.2. HP Metal Jet technology ................................................................................. 10
2.3. Scope ........................................................................................................... 12
3. STATE OF THE ART ______________________________________ 13

3.1. Machine development process..................................................................... 13
3.1.1. Traditional sequential engineering .................................................................. 13
3.1.2. Systems engineering and concurrent design .................................................. 15
3.2. Virtual Commissioning.................................................................................. 16
3.2.1. Hardware in the loop ....................................................................................... 17
3.2.2. Software in the loop ........................................................................................ 17
3.2.3. Virtual commissioning software ....................................................................... 18
4. SYSTEM DESCRIPTION ___________________________________ 20

4.1. Functionality ................................................................................................. 20
4.2. Hardware architecture .................................................................................. 22
4.2.1. Cartesian robot ............................................................................................... 22
4.2.2. Conveyor......................................................................................................... 24
4.2.3. Camera ........................................................................................................... 25
4.2.4. Controller ........................................................................................................ 26
4.3. Simulation environment ................................................................................ 26
4.3.1. Time management .......................................................................................... 29
4.3.1.1. Time slices ........................................................................................... 29
4.3.1.2. Operation modes .................................................................................. 29
5. DEVELOPMENT OF THE DIGITAL TWIN ______________________ 33

5.1. Mechatronic simulation ................................................................................. 33
5.1.1. Basic physics .................................................................................................. 33
5.1.1.1. Rigid bodies ......................................................................................... 33
5.1.1.2. Collision bodies .................................................................................... 35

5.1.2. Joints and constrains ...................................................................................... 35
Pg. 6 Report
5.1.3. Couplers ..........................................................................................................36

5.1.4. Sensors............................................................................................................37
5.1.5. Actuators..........................................................................................................38
5.2. User program ............................................................................................... 38
5.3. Controller simulation .................................................................................... 41
5.4. Data exchange ............................................................................................. 41
5.5. Drivers behaviour simulation........................................................................ 42
5.5.1. Drive X .............................................................................................................43
5.5.2. Drive Y and Z ...................................................................................................45
6. VALIDATION OF THE SIMULATION __________________________ 49

6.1. Operation of the simulation .......................................................................... 49
6.2. X-axis motion ............................................................................................... 50
6.3. Y-axis motion ............................................................................................... 54
6.4. Z-axis motion ............................................................................................... 58
6.5. Discussion.................................................................................................... 62
7. ECONOMIC ANALYSIS ____________________________________ 64

8. ENVIRONMENTAL IMPACT ________________________________ 67
CONCLUSION _______________________________________________ 69
ACKNOWLEDGEMENTS _______________________________________ 71
BIBLIOGRAPHY ______________________________________________ 72
Citations................................................................................................................. 72
Complementary bibliography................................................................................. 73
1. Glossary
MCD: Mechatronics Concept Designer
PLC: Programmable Logic Controller
HP: HP Inc.
VC: Virtual Commissioning
HMI: Human-Machine Interface
PC: Personal Computer
TO: Technology Object

Pg. 8 Report
2. Introduction
2.1. Objectives
Machine manufacturers are facing a higher demand to reduce the time-to-market of their
new systems. At the same time, customers expect to receive high-quality, customizable
products at a reasonable price. This creates a need to re-think the development cycle to
continue being competitive in the market.
In the industrial machinery business, equipment must be productive, flexible, robust and
inexpensive. Modularity and reusability also become very important to adapt to customers
specific demands. Meeting these requirements while developing new technologies like
additive manufacturing, is a great business challenge. The development process is
complex and with high degrees of uncertainty. It needs a lot of innovation and multiple
disciplines collaborating, agreeing on the requirements and integrating their work. It also
requires ways to shorten or parallelize design activities to reduce lead time.
In this environment, the industry needs new development methods and tools that help
them deal with complexity, be more efficient, and reduce costs. Many tools are already
available in the market. These software tools enable building prototypes and testing
machine control programs in virtual models that emulate physical systems. These models
allow sharing system functionality and constraints, analysing design concepts and validate
control programs to make informed decisions before any hardware is built. At the same
time, they enable an earlier integration of the different disciplines and a parallelization of
design activities.
Having said that, the use of these tools is not yet widespread across the industry, as it
requires an initial investment on licenses and skilled professionals with the know-how to
build and operate the machine models.
In this context, this project aims to develop a case study where a system is developed using a
model-based approach. The system is a prototype built to automatically pick 3D printed parts
and place them in a packaging box. The 3D printed parts can have any size and shape within
the machine limits. To perform this operation, pick and place algorithms have to be developed,
validated and optimized. In this scenario, it is very useful to have a tool to verify the correct
operation of the code with a 3D visualization of the movement sequence of the machine in a
virtual environment, limiting the risk of damaging physical equipment. Thus, in this work, a
model of the system is going to be developed to validate the system control logic before testing
it in the physical hardware.
This project also aims to analyse and discuss further the advantages and implications of
model-based development from an organizational perspective. This will be done through a
literature review and the observations made during the development of the case study.
2.2. Context
2.2.1. Usage of programmable logic controllers
HP Inc. develops 3D printing and digital manufacturing solutions for industrial applications. As
most of the industrial machinery, some of these solutions use Programmable Logic Controllers
(PLC) to manage the logic of the system.
PLC are off-the-shelf electronics. One of the reasons to use PLCs in industrial machines is
that they are expected to be reliable in harsh electrical and physical environments. They also
comply with industry standards and are certified by standard organizations, like CE or UL.
Besides, it generally takes less time to program a PLC than an embedded system, and they
are known to be modular and easily replaceable if a specific module fails. Although PLCs have
a high cost compared to custom electronics, it usually makes sense to use them in low-volume
industrial projects where time-to-market is very important.
The use of PLCs comes with a toolset that not only includes programs for software
development but also tools that connect PLC code with mechatronic designs into a virtual
model, known as the digital twin. Developing machines using this toolset can bring several
benefits to machine manufacturers. Some of them are:
1. The machine can be commissioned before real hardware is assembled, enabling the
parallelization of development activities and reducing the time to market.
2. Errors are more likely to be detected in an early stage of the development process and
without damaging any physical equipment.
3. Software debugging and testing can be done without using hardware resources.
4. The software can be developed by multiple users from multiple sites.
5. It simplifies the outsourcing of some subsystems, in cases in which it is believed that
other parties can develop those systems more quickly, or cheaper.
6. It makes easier to define and communicate the functionality of the system between
engineering disciplines.
7. It assists in getting aligned with customers' requirements.
As the prototype in this project has a low-volume industrial application and a short development
cycle, a PLC has been chosen to control it. The control program will be tested against a virtual
Pg. 10 Report
model created with some of the PLC simulation tools.
2.2.2. HP Metal Jet technology
The present project has been developed within HP Inc. at the Research and Development
department of the 3D Printing division. HP has developed two additive manufacturing
technologies: HP Multi Jet Fusion for plastics and HP Metal Jet for metals. This project has
been developed inside the HP Metal Jet program.
HP Metal Jet is an additive manufacturing technology for mass-production. Thus, it is important

to automate certain parts of the manufacturing process to reduce time and labour costs.
Automated processes also increase repeatability and yield by minimizing human errors. Pick
and place operations, like the one performed by the machine in this project, are a good
example of standard tasks that can be automatized.
HP Metal Jet technology falls into the binder jetting category inside additive manufacturing
technologies [13] according to the classification in Figure 1. The principle of binder jetting is to
build a part by glueing powder particles together employing a binding agent.
Figure 1: Additive Manufacturing Technologies [13].

The process to form a metal part using HP Metal Jet technology is shown in Figure 2 and
described below.
Figure 2: Schematic of HP Metal Jet additive manufacturing process [12].
Firstly, a build unit is loaded with powder and introduced to the printer. Then, the process
continues as follows:
1. Spread powder: the process begins with a recoater spreading a layer of metal powder.
2. Print agent: printheads jet the binding agent at precise locations onto the powder bed
to define the geometry of the parts.
3. Evaporation: the powder bed is heated throughout the printing process to evaporate
the liquid components of the binding agent.
4. Retract the bed, print next layer: the powder bed is retracted the thickness of the printed
powder layer, and the process repeats until the build is completed.
5. Cure the bed: the powder bed with its printed parts is heated to complete the
evaporation of liquid components from the binding agent and to cure the polymers.
6. Decaking: the powder bed is cooled and then the parts can be extracted. Decaking is
the process of removing loose powder from the surfaces of the parts. Once the parts
have been decaked and extracted, the remaining powder can be processed and
reused.
7. Sintering: the parts are moved into a furnace. At sintering temperatures, a diffusion
process takes places at the surface of the metal particles. Also, the polymer from the
binding agent decomposes. As a result, the metal particles are bonded together, and
the final parts are formed.
8. Finishing: the parts may go through machining and polishing processes to meet
dimensional and surface finish requirements.
Pg. 12 Report
Once the parts are finished, they are packed and shipped out to customers. The machine in
this project is meant to be used in this final step to prepare the parts to be sent.
2.3. Scope
In this project, a system will be developed using a model-based approach. A virtual model of
the system is going to be developed to validate the control program in a simulated
environment. This will allow early detection of errors in the control software and thus reduced
risk to cause any physical damage when testing the code in the real machine. It will also allow
several developers to work on the control program simultaneously without having access to
the prototype. In addition, building the simulation will provide a deeper understanding of the
system.
The system is a prototype built to automatically pick 3D printed parts from a tray and place
them in a packaging box. The 3D printed parts can have any size and shape within the machine
limits. The mechanical solution of the system has already been designed. It consists of a
gripper mounted on a 2-axis cartesian robot and a conveyor that transports the packaging box
and the tray with the printed parts.
The machine model must be able to understand the position setpoints introduced by the user
and provide a physics-based representation of the machine motion. The following machine
characteristics will be simulated to achieve this goal:
- The physical properties and kinematics of all the machine components with motion.
That is the rigid bodies mass and inertia, the joints, the sensors and the actuators.
- The behaviour of the drives (motor controllers).
- The data exchange between the controller and the peripheries (motors, sensors and
drives).
- The controller governing the machine.
The simulation is going to be controlled by the same program that governs the real machine.
However, the control program of the conveyor was not ready when starting this project. Thus,
a program will be developed by the author of this work. This will be done to be able to test the
simulation of the complete machine.
The virtual model will be validated by a 3D visualization of the machine motion as well as an
analysis of the motion graphs of several given movement sequences. The project also includes
an analysis of the costs and the environmental impact of this work. Lastly, some conclusions
are derived from the experience of creating the virtual model of a machine during its
development, determining if this can help to build new hardware more efficiently, with better
quality and less cost.
3. State of the art
3.1. Machine development process
In the industry, all projects aim for achieving the maximum quality and the minimum cost within
a tight schedule. While this is broadly accepted, it is very hard to achieve these three objectives
simultaneously. Pretty often, project managers find themselves forced to sacrifice one or two
of the goals to deliver a product that meets the specifications, the price, and the timeline
required by the customers. It is quite common to encounter situations where companies go
over the budget or are required to reduce the project scope to deliver a product on time. Other
times they must delay the schedule to achieve the desired product value at a reasonable cost.
The root cause of these problems is linked to having high degrees of uncertainty. Uncertainty
in content (for many projects the work has never been done before and the customer may not
have fully determined what they want), in vendor performance and internal skill sets [4].
To reduce project timelines, machine manufacturers are introducing innovations in internal

design processes. One of the approaches is the introduction of systems engineering and
concurrent design.
3.1.1. Traditional sequential engineering
Machines are nearly always developed with a combination of disciplines: mechanical, fluids,
thermal, electrical and software. In the traditional sequential engineering process, one
discipline comes after another, as illustrated in Figure 3 and Figure 4.
Figure 3: Traditional machine development process [8].
At the start of a project, external customers or internal project managers formulate the
requirements, usually quite roughly. Requirements are often only considered during the
scoping phase of the process and not revised afterwards. Then, conceptual designs are
formulated, being mainly mechanically driven. Moving to the detailed design, mechanical
engineers are the ones who start first. Their work results in the CAD of the machine, the
drawings of the components and the manufacturing of all the parts. While the machine is being
manufactured and build, mechanical engineers deliver the list of sensors and actuators to
electrical engineers, who make the wiring diagrams and assign the inputs and outputs of the
Pg. 14 Report
hardware. Only after the IO assignment is finished, the software developers begin to program
the software that controls the machine. Therefore, software engineering tends to be the last
step within the development process [3] and it is usually carried out during the commissioning
phase.
Figure 4: Traditional, sequential engineering process [7].
This sequential way of engineering comes with some drawbacks, especially in the age of
industry 4.0, when the software content of a machine continuously increases [7], as shown in
Figure 5.
Figure 5: Evolution of software content of machines [7].
Some of these drawbacks are:
- Longer lead time: the commissioning of the assembled machines can take up to 15-
25% of the total duration of the project according to [3]. If these activities are done
sequentially and with the machine stopped, they can potentially cause a delay in the
delivery date.
- Late design error identification: if control programs are debugged in the real machine,
unexpected design errors and derived machine damage or people injury are likely to
arise, as the software is not mature enough.
- Requirements mismatch: since customers' requirements are only revised at the

beginning of the project, the focus on the detailed design can lead to a result that
differs from the customers' expectations. Besides, when the functionality definition
focuses only on the mechanical side of the machine, integration issues might appear
in the later stages of development, affecting the quality of the final product or
introducing unexpected outlays.
3.1.2. Systems engineering and concurrent design
New developing methods such as systems engineering and concurrent design have been
spreading across the industry to improve the traditional process.
Systems engineering is a field that focuses on how to effectively design and manage complex
systems with an interdisciplinary and holistic approach. The development process starts with
a detailed definition of the requirements of the system. This can be done through a conceptual
system model where the various engineering disciplines can work together to define the
functions and processes in every given operating mode of the machine. As soon as these
requirements are stated, each discipline can use its tools to perform the detailed design. Then
again, simulation models are used to check the proper functioning of the software and integrate
and validate the functionality of the machine. This allows the detection of errors long before
any physical prototype is built and substantial time savings. This process is illustrated in Figure
6 and Figure 7.
Figure 6: Today’s machine development process [8].
Systems engineering also involves customers in the design process and ensures that the
system developed is viable throughout its life [8]. Requirements are tracked through the design
process and not only in the beginning. The result is a higher quality product developed more
efficiently.
Pg. 16 Report
Figure 7: Engineered systems development process [7].
3.2. Virtual Commissioning
In traditional machine development, software programs are tested, in the first place, in the
control software development tools themselves, without real feedback of the process
behaviour. Then, with the system already assembled and connected, the commissioning is
carried out. However, with the growth of the computational capacity of computer equipment
and the irruption of tools for emulating machinery and processes, we are now able to test the
software before the real commissioning. This process is called virtual commissioning (VC) [1].
Virtual commissioning consists of using simulation tools to emulate a physical system

(mechanical components, sensors, actuators, motor drives, valves, pumps, controllers, HMI,
etc.) in a virtual environment during the development phase. This allows coding of the system
before the physical hardware is installed. It also gives the developer the option to confirm the
correct operation of the code as well as the physical movement and feedback from the
hardware before any equipment is built. Virtual commissioning also allows for easy
reconfiguration of an existing system, where process, software or hardware changes can be
made in the digital model [2].
The real-time digital representation of the system is known as the digital twin. Figure 8 provides
a schematic of the elements that can be emulated through simulation tools (events, flow,
behaviour, geometry and kinematics) and the hardware than can be tested against these
simulations (RC, HMI, PLC, MES and ERP). In this project, a PLC program is validated against
a simulation of events, behaviour, geometry and kinematics.
Figure 8: Virtual commissioning of control systems [5].
Virtual commissioning can be performed in two configurations: hardware-in-the-loop and

software-in-the-loop. They are described in the following two chapters.
3.2.1. Hardware in the loop
Hardware-in-the-loop (HIL) configuration requires having a virtual model of the process or

machine and a real controller. The real PLC or control hardware is connected to the model. It
might involve some additional hardware and Fieldbus devices apart from the controller. This
configuration is generally more expensive, but the testing environment is very similar to the
final one, as the software is directly tested on the controller. Figure 9 provides a schematic of
a HIL configuration.
model
ieldbus Controller User program
device
Figure 9: Hardware-in-the-loop (HIL) configuration.
3.2.2. Software in the loop
In a software-in-the-loop (SIL) configuration, the model of the system is controlled by a virtual

controller. This provides a very flexible solution as the only hardware needed is a personal
computer. The software can be developed simultaneously by different engineers based around
the world against the same system model. However, every simulation software brings a license
cost associated with it, which must be considered when evaluating emulation benefits. Figure
10 provides a schematic of a SIL configuration.
Pg. 18 Report
model
Virtual User program
controller
Figure 10: Software-in-the-loop (SIL) configuration.
3.2.3. Virtual commissioning software
There are numerous tools for emulating industrial systems and processes. ABB Robot Studio,
Excelgo Xperior, Simumatik3D or Siemens NX Mechatronics Concept Designer are tools to
model the geometry and kinematics of the system [1].
Other tools such as Siemens SIMIT or Simulink can be used for modelling machines and
processes in applications such as water treatment plants. SIMIT also simulates the behaviour
of devices like sensors or drives for motors, valves, and pumps. However, it does not model
the geometry or kinematics of the system. Similar tools like Simscape and Simcenter Amesim
enable to create simulations with libraries of ready-to-use models of physical systems (electric
motors, refrigerator systems, hydraulic actuators, etc.).
Tecnomatix Process Simulate is a tool that provides the possibility of simulating the motion of
a robotic cell within a production line and detecting collisions between a robot and the
production machine or between two robots.
Figure 11 provides an overview of the simulation levels of Siemens virtual commissioning

software. On the automation level, PLCSIM Advanced and WinCC are the Siemens tools that
simulate the PLC and the human-machine interface (HMI), where the user can interact with
the system and introduce commands to control it.
Figure 11: Simulation levels of Siemens virtual commissioning software [8].

All these tools make the creation and adjustment of the model much easier and faster than if
it had to be built from scratch without employing reusable components.
Siemens virtual commissioning programs provide a fast, easy and reliable integration with
Siemens PLCs or their simulation in PLCSIM Advanced. As our machine is controlled by a
Siemens PLC, it makes sense to use the Siemens toolkit. Therefore, NX Mechatronics
Concept Designer, SIMIT and PLCSIM Advanced will be used to build the digital twin of our
system and perform the virtual commissioning. The purpose and characteristics of these tools
are explained with more detail in chapter 4.3.
Pg. 20 Report
4. System description
4.1. Functionality
Virtual commissioning can be performed in a wide variety of systems. This project shows an
example of one of many applications.
The purpose of the machine in this work is to perform a pick and place operation to transport
3D parts manufactured using the HP Metal Jet technology, explained in chapter 2.2.2. HP
Metal Jet parts can have any size and geometry within the limits of the printing bucket. Figure
12 shows sample parts produced with this technology.
Figure 12: Sample 316L stainless steel parts created using HP Metal Jet technology [12]
Throughout the production process, the 3D printed parts are moved from one process to
another on a tray. When they reach the end of the manufacturing line, the parts are placed in
packaging boxes before shipping them out to customers. In a mass-production environment,
it is key to automate this task to reduce time, labour costs and human errors. That is why a
system is being developed in HP to pick and place the final 3D printed parts automatically.
The system must be able to identify the location and geometry of the parts within the tray that
carries them. At the same time, it must be able to position a gripper at any height in the range
of 0 to 200 mm. Finally, it must be capable of reaching any point in the XY dimensions of the
tray and the packaging box where parts will be placed. Figure 13 shows a schematic of the
system architecture.
Parts tray Packaging box
Figure 13: Top view of a schematic representation of the system architecture.
The gripper is positioned using a 3-axes actuator, which is described in chapter 4.2. The
geometry and location of the parts are acquired with a 3D camera at the beginning of the
process. Having this setup, pick and place algorithms have to be developed to move any set
of 3D printed parts to the respective packaging box. Those algorithms process the images
coming from the 3D camera, get the coordinates of every printed part of the tray, define the
movement sequence and send it to the 3-axis actuator. This is done in a PC. The PC then
communicates with a PLC that controls the operation of the 3 axes of the actuator. Figure 14
provides a schematic of the interaction between the hardware elements.
PC
PLC
Camera
Figure 14: Front view of a schematic representation of the system architecture.
The program that controls the simulation is the same that governs the physical machine.
Before starting any movement sequence, it requires homing the 3 axes. Then, it starts reading
cyclically the position setpoints sent by the program running in the PC. In this project, all the
hardware of the system has been simulated except for the camera and the PC. Thus, the
position setpoints of the 3 axes are introduced manually. The virtual model provides a
Pg. 22 Report
visualization of the movement sequence of the system. The simulation tools used are
described in chapter 4.3.
4.2. Hardware architecture
The mechanical solution for the pick and place process is a 3 axes actuator system composed
by a 2-axis cartesian robot and a conveyor belt. The cartesian robot moves in the y and z axes,
as shown in Figure 14. The z-axis carries the gripper to pick and place the 3D printed parts.
At the beginning of the process, the 3D camera captures the location and geometry of the
printed parts, which are transported on a tray. Then, the pick and place algorithms are
calculated in a PC and the position setpoints are sent to the PLC that controls the machine.
The tray is then moved under the cartesian robot by the conveyor belt and the cartesian robot
is displaced to pick one part from the tray. The conveyor belt also transports the box where the
3D printer parts will be packed. Then, the conveyor belt moves until a line of slots of the
packaging box is under the cartesian mechanism. After that, the z and y axes move again to
place the carried part inside the box. Finally, the conveyor belt moves the tray under the
cartesian robot to start the process again for the next part.
The characteristics of each piece of equipment used in this solution are described in the
following chapters. The datasheets of the simulated devices are also available in Appendix B.
Some of this information will be later used as an input for the simulation.
4.2.1. Cartesian robot
The cartesian robot consists of two axes. The y-axis provides horizontal movement to the
carriage that transports the z-axis. The z-axis provides vertical movement to the gripper, which
is not included in this simulation as it is not selected yet. Figure 15 shows the cartesian robot
in the home position.
Figure 15: 2-axis cartesian robot at home position

The mechanical characteristics of the two axes are the following:
- Y-axis: the stroke of the mechanism is 297,33 mm, although it is limited by software to
280 mm. The transmission is done through a timing belt and pulley mechanism, which
is shown in Figure 16. The linear travel per motor revolution is 70 mm. The gear ratio
is i=1. The pulley shafts are mounted on ball bearings and an elastic coupling transmits
the torque from the motor shaft to the pulley shaft.
Figure 16: Section of the Y-axis mechanism
- Z-axis: the stroke of the mechanism is 350 mm, although it is limited by software to 240
mm to avoid a collision between the gripper and the tray carrying the parts. The
transmission is done directly from the motor axis via a rack and pinion mechanism. The
linear travel per motor revolution is 72,25 mm. An elastic coupling transmits the torque
from the motor shaft to the gear shaft. Figure 17 provides a close view of the z-axis
mechanism.
Figure 17: Detail of the Z-axis mechanism.
The y and z axes of the cartesian mechanism are an Igus S.L. module with reference DLE-
LG-0158. The power in both axes is provided by an integrated stepper motor from Schneider
Electric, the Lexium MDrive with reference LMHCE517C. It has a built-in controller and an
Pg. 24 Report
incremental magnetic encoder. The drive is integrated with the motor. Each motor turn is
equivalent to 51200 micro-steps.
The home positions are detected by two limit switches from Telemecanique (reference
XCMD4102L5EX).
4.2.2. Conveyor
The x-axis of the mechanism is a conveyor belt, shown in Figure 18. The belt carries the tray
with the 3D printed parts and the packaging box. The tray and the box are attached to the belt
by friction. Thus, the belt has been selected to prevent slip.
The effective length of the conveyor is 1650 mm. It is powered by a brushed DC motor form
Dunkermotoren, reference GR 63X55 40V. It has a nominal speed of 3450 min-1 and a no-load
speed of 3600 min-1. The motor comes with a brake and an incremental encoder of the same
provider, with reference E-90 and RE-30 respectively.
The drive used to control this motor is the SIMATIC Micro-Drive PDC from Siemens.
Figure 18: Top view of the conveyor.
The motor rotation is converted into the linear movement of the belt in three steps, illustrated
in Figure 19. The first step of the transmission is a planetary gearbox (Dunkermotoren
reference PLG52) with i=1/288 gear ratio. The second stage is a timing belt and pulley with
i=1 gear ratio that transfers the power to a second axis. The third step is another mechanism
consisting of a timing belt, a timing pulley and several idle pulleys as belt tensioners on each
side of the conveyor. This system converts the rotary movement of the second axis to the
linear movement of the conveyor belt. The linear travel per revolution is 150 mm. Figure 20
shows a lateral view of the conveyor transmission system.
Figure 19: Transmission steps of the conveyor belt actuation.
Figure 20: Lateral view of the conveyor transmission mechanism.
The home position is detected by a limit switch from Telemecanique (reference

XCMD4102L5EX) that is activated by the tray that carries the 3D printed parts. Figure 21
shows the location of the X-axis home sensor.
Figure 21: Location of the X-axis home sensor.
4.2.3. Camera
The camera used to capture the height and location of the printed parts is the Intel®
RealSense™ D415 depth camera. It is a low-cost camera designed for easy set-up and
portability, that allows building applications using sample code. Despite its small form factor, it
integrates two depth sensors (IR cameras), an RGB sensor, an infrared projector and a low
power vision processor. It has a standard depth field of view (FOV) of 65°±2°x 40°±1° x 72°±2°
Pg. 26 Report
that is well suited for accurate computer vision. It supports indoor and outdoor lighting
conditions. The distance to objects is obtained through triangulation by comparing the
distances measured by the two depth sensors.
4.2.4. Controller
The controller used in this system is a programmable logic controller (PLC). A PLC is an
industrial computer that is programmed to monitor input signals, perform logical operations,
and trigger specific output signals. In most cases, PLCs are modular, which means you can
install I/O modules in a plug-and-play manner. The PLC used for this application is the CPU
1512SP-1 PN provided by Siemens. Apart from the CPU, it also includes a power supply, a
rack to plug in the modules and the input and output cards.
Since a PLC is a dedicated controller, it will only process one program repeatedly. One cycle
through the program is called a scan time. It involves reading the inputs from the other
modules, executing the logic based on these inputs, and then update the outputs accordingly.
The scan time happens in the range of 1ms [14].
Siemens PLCs mainly use the Profinet communication protocol to communicate with other
devices like drives or HMIs. Profinet is an Ethernet-based industrial communications protocol
that works at 100 megabits per second. Due to its high-speed operation and the response time
of less than 1ms, Profinet is suitable for high-speed applications. Profinet cables have robust
shielding to function well in harsh environments, which make them appropriate for PLC
applications.
4.3. Simulation environment
As explained in chapter 3.2.3, this project uses the Siemens virtual commissioning toolset to
build the model of the mechatronic system. Figure 22 shows the functions and interactions
between the four programs composing the toolkit:
- TIA Portal
- PLCSIM Advanced
- SIMIT
- NX Mechatronics Concept Designer
Figure 22: Interactions between the software of the Siemens virtual commissioning toolkit [9].
TIA (Total Integrated Automation) Portal is the software used to program the code of the PLC,
the so-called user program. A TIA Portal project also includes the hardware configuration: the
controllers, screens, periphery, actuators and communication.
Once the TIA Portal project is finished, it can be tested and validated using PLCSIM Advanced,
which simulates the controller (the PLC). PLCSIM Advanced also takes care of the
communication and synchronisation with external applications, like MCD or SIMIT.
NX Mechatronics Concept Designer (MCD) is the tool used to build the physical and kinematic
model. The user can introduce the physics and kinematics of the machine components over a
previously existing 3D model (CAD) of the machine. That includes defining the mass and
inertia of the rigid bodies, the collision bodies, the kinematic joints and constraints, the sensors
and the actuators. With this model, the engineer can generate and validate the operation
sequence and motion of the machine or the load curves of the actuators. It can also be used
to detect collisions and mechanical interferences happening during the machine operation.
The aim of this simulation is not to conduct a detailed analysis and validation of the specific
characteristics of individual components (vibration, thermal, frequency response, or stress),
but to simulate the overall machine behaviour [6].
MCD is also used by mechanical engineers to create and validate machine concepts in the
early stages of development. Likewise, the data of their models can be used by electrical
designers to select sensors and actuators. At the same time, automation engineers can use
the cams and operation sequence information from the same model for software development
[11].
SIMIT emulates the behaviour of devices like sensors or drives for motors, valves, and pumps.
It can also simulate process variables like temperature, pressure or flow. The models are built
using standard libraries or user-made components. SIMIT is also used to communicate and
Pg. 28 Report
synchronize the mechatronics model in MCD with the controller model in PLCSIM Advanced.
It automatically detects the signals coming from both programs, which are transformed and
adapted to be understood by both tools.
SIMIT also provides a user interface with graphics and controls that can be used to train
operators, monitor or control signals of the machine and force errors to test the software.
Figure 23: Data exchange between PLC, SIMIT and MCD [9].
Figure 23 explains the interaction and signal exchange between PLCSIM Advanced, SIMIT
and MCD:
- SIMIT receives the outputs from the PLC.

- SIMIT sends setpoints (e.g. speed setpoints, position setpoints, binary signals) to
MCD.
- MCD simulates the mechanical components movements based on the defined
degrees of freedom and received setpoints.
- MCD sends the calculated actual values (e.g. speed actual values, position actual
values, binary signals) to SIMIT.
- SIMIT calculates the behaviour of the machine using the drive logic defined.
- SIMIT sends the inputs to the PLC.
4.3.1. Time management
4.3.1.1. Time slices
In a simulation, the cycle time is the time frame in which the calculations of the models have
to be executed and the data has to be exchanged. In SIMIT, the cycle times are set using one
of the eight available time slices in the project, as shown in Figure 24. The minimum cycle time
that can be set is 1ms.
Figure 24: Project times and operating modes definition in SIMIT.
Then, each coupling, chart or individual component can be assigned to a different time slice.
For couplings, the time slice is the time rate at which they exchange data. For charts and
individual components, it corresponds to the assigned calculation time.
4.3.1.2. Operation modes
SIMIT has 3 modes of operation when it comes to time management: asynchronous,

synchronous and bus synchronous. The operating mode is selected in SIMIT in the window
shown in Figure 24.
Asynchronous operating mode
In asynchronous mode, the calculation of the model and the exchange of signals in the
coupling are triggered cyclically. That means that it is time controlled. The model calculation
continues even if the signal exchange with one coupling is blocked by a communication
problem or not calculated within the allocated time. However, the processing stops if one or
more machine models in SIMIT are not calculated within its allocated time. In that case, it will
only continue when all SIMIT models meet their cycles again. There is a time offset between
the data communication of couplings and the model calculation, as shown in Figure 25.
Pg. 30 Report
Figure 25: Time offset between the signal exchange and the model calculation in asynchronous
operation mode [10].
Figure 26 illustrates an example of a simulation performed in asynchronous mode. The system

has a coupling with a simulated controller in PLCSIM Advanced. The simulation model is
calculated in SIMIT. Both the controller and the simulation model have a cycle time of 100ms
and they are triggered with a time offset between them.
Figure 26: Example of the operation of an asynchronous simulation [10].
The numbers in Figure 26 represent the steps of the simulation:
1. The controller has finished its calculation cycle. Signals are available at the outputs of
the controller.
2. The output signals of the controller are applied to the SIMIT coupling memory. Here
they are available for the calculation of the model.
3. The calculation of the model is triggered.
4. End of the model calculation cycle. The calculated input signals for the controller are
now available in the coupling memory.
5. With the next clock pulse of the coupling, the input signals are transferred to the
controller.
6. If the controller only evaluates its inputs at the start of a cycle, the transferred input
signals are only used in the control program now. This means the reaction of the
simulation does not take place until two control cycles later.
Synchronous operating mode
In synchronous operating mode, the simulation calculation and the signal exchange are done
in a specified sequence. One action only starts after completing the previous one, as shown in
Figure 27. This results in better response times for short cycle times. However, each module
and each coupling can block the entire simulation execution.
Figure 27: Simulation execution in synchronous mode [10].
Bus synchronous operating mode
In bus synchronous operating mode, all the components involved in the simulation have the
same execution cycle. As shown in Figure 28, the real plant cycle time always matches the
simulation cycle time, no matter how long it takes to calculate a simulation step.
Pg. 32 Report
Figure 28: Correspondence of the simulation cycle times with the real-plant cycle times in bus
synchronous operation mode [10].
The top part of Figure 28 shows how long PLCSIM Advanced, SIMIT and MCD take for their
calculations. Each cycle starts only when all the calculations are done, and the results have
been exchanged through the couplings. The bottom part of the figure represents the timing of
the CPU, drive and axis in the real plant.
Bus synchronous operating mode is suitable for applications with real-time requirements and
it is possible with PLCSIM Advanced and MCD. That is why it has been chosen for this project.
In bus synchronous operating mode, a time slice of the project must correspond to the cycle
time.
5. Development of the digital twin

In this chapter, the models created to simulate each component of the system and the
configuration of the simulation are explained.
5.1. Mechatronic simulation
The mechatronic simulation is done by giving physical properties and kinematic constraints to
the 3D model of our machine. This is done in the NX Mechatronics Concept Designer (MCD).
5.1.1. Basic physics
The basic physics is the definition of the rigid bodies and the collision bodies of the system.
5.1.1.1. Rigid bodies
In MCD, rigid bodies are components that can move. They have mass, inertia, translational
and angular velocity, position and orientation. They respond to forces like gravity. In general,
all the objects that move together are part of the same rigid body. Stationary objects do not
have to be defined as rigid bodies and are part of the background. This prevents the need to
simulate the structure of the machine, for example.
To create a rigid body, the user has to select all the objects to be included. Then, the mass
and inertia of the body are automatically calculated based on the material properties of each
component, which have to be previously available in the CAD model. Otherwise, the user can
also introduce the mass and inertia properties manually. Figure 29 shows the dialogue box in
MCD to create a rigid body.
Figure 29: Dialogue box to create the Y carriage rigid body.

Pg. 34 Report
The rigid bodies that have been defined in MCD are:
a) The tray carrying the printed parts

b) The Z carriage
c) The Y carriage
d) The X motor shaft
e) The X gearbox output shaft
f) The X driving shaft
The 3D model of the gearbox is a single component. Hence, the rigid body of X gearbox output
shaft has been represented by the timing pulley linked to it. Figure 30 illustrates the six rigid
bodies of the mechatronic simulation.
(a) (b) (c)
(d) (e) (f)
Figure 30: Rigid bodies in MCD. (a) Tray, (b) Z carriage, (c) Y carriage, (d) X motor shaft, (e) X
gearbox output shaft, (f) X driving shaft.
Note that the movement of the z and y axes will be applied directly to the carriages, while in
the x-axis, the whole transmission is simulated in MCD. The transmission ratios of the y and z
axes are introduced in SIMIT as part of the drives’ behaviour model. Those are two alternatives
to perform the kinematics simulation. The first one reduces the computational needs in MCD,
while the second one provides more accurate results, as it takes into account the physical
properties of the transmission components. Furthermore, when trying to simulate the z and y
axes transmission in MCD, the simulation has an unexpected behaviour. This behaviour has
been reported to Siemens and it has been classified as a bug in the software. That is why the
option to simulate the transmission of the z and y axes in SIMIT has been chosen.
5.1.1.2. Collision bodies
Collision bodies define how elements collide with other elements that also have a collision
body. Elements without a collision body pass through other objects. Collision bodies are not
rigid bodies, but they can be part of one. MCD calculates collisions using simplified collision
shapes that encapsulate elements [15]. Increasing the geometric accuracy of the collision
shapes may reduce simulation performance and cause simulation failures like objects passing
through, sticking together or jittering. Table 1 shows the collision shapes available in MCD and
their geometric accuracy, reliability and simulation performance.
Type Geometric accuracy Reliability Simulation performance
Low High High

Box
Low High High

Sphere
Cylinder Low High High
Low High High

Capsule
Medium High Medium

Convex
Medium High Medium

Multi Convex
High Low Low

Mesh
Table 1: Collision shapes in MCD and their geometric accuracy, reliability and simulation performance
[15].
The user can define the collision material and set its dynamic friction, static friction, rolling
friction and restitution coefficients.
In this project, the surfaces of the rigid bodies that collide with the home sensors are defined
as collision bodies with a box shape. The home sensors themselves are defined as collision
sensors, which are explained in chapter 5.1.4.
5.1.2. Joints and constrains
Joints and constraints set the degrees of freedom of the rigid bodies. In this project, only hinge
and sliding joints are needed. Hinge joints allow one rotational degree of freedom along an
axis between two rigid bodies. Slide joints allow one translational degree of freedom between
two rigid bodies along a vector.
Pg. 36 Report
The joints that have been set in MCD are:
1. A hinge joint between the X motor shaft and the background.

2. A hinge joint between the X gearbox output shaft and the background.
3. A hinge joint between the X driving shaft and the background.
4. A sliding joint between the tray and the background.
5. A sliding joint between the Y carriage and the background.
6. A sliding joint between the Z carriage and the Y carriage.
Figure 31 shows the coordinate system of the machine and the joints vectors numbered as the
list above. In the hinge joints, the rotation axis of the joint coincides with the axis of each shaft
in the Y positive direction. The joint rotation anchor point is the centre of each shaft. In the
sliding joints, the sliding vectors correspond to the directions of the coordinate system.
Figure 31: Machine coordinate system and joints vectors.
5.1.3. Couplers
Couplers are used to simulate the coordinated motion of multiple axes. In this project, two
gears and one mechanical cam coupler are used to simulate the transmission steps of the x-
axis. A mechanical cam is a coupler that connects the motion of two axes and forces them to
maintain a relationship determined by a motion profile. The motion’s reaction force is
transferred back to the master axis. A gear coupler forces two rotating axes to maintain a
constant turn ratio [15].
The first gear coupler represents the gearbox transmission. The hinge joint of the X motor shaft
(1) is set as the master and the hinge joint of the X gearbox output shaft (2) as the slave. The
gear ratio is 288:1.
The second gear coupler simulates the belt transmission between the gearbox output shaft
and the second shaft. The master is the hinge joint of the X gearbox output shaft (2) and the
slave is the hinge joint of the X driving shaft (3). The gear ratio is 1:1. As this transmission uses
a toothed belt, the option to allow some slip in the coupler is not selected.
The mechanical cam simulates the transmission of the rotary movement of the driving shaft to
the linear movement of the tray. The master is the hinge joint of the X driving shaft (3) and the
slave is the sliding joint of the tray (4). The linear travel per revolution ratio is set using a motion
profile shown in Figure 32. The ratio is 150 mm/rev. As this transmission uses a toothed belt,
the option to allow some slip in the coupler is not selected.
Figure 32: Motion profile defining the motion of the slave axis relative to the motion of the master axis.
5.1.4. Sensors
In this simulation, there are three collision sensors. They correspond to the roller plungers of
each limit switch used as a home sensor. Collision sensors are very similar to collision bodies,
but they also provide feedback about the interaction of rigid bodies during the simulation. The
status of the collision sensor can be used to trigger mechatronic operations like starting or
stopping operations, changing the speed of an actuator, creating new objects or detecting their
position, among others. In this case, the status of the collision sensors is sent to SIMIT so that
the motor drives and the PLC can detect when the axes are at the home position.
Pg. 38 Report
5.1.5. Actuators
As there are three motors in the machine, three actuators are needed. The actuators can be
electrical, pneumatic or hydraulic. In the present case, all the actuators are electrical. The y
and z motors have been defined as position control actuators and the x motor as a speed
control actuator.
Position control actuators use the position as the driving parameter, while speed actuators use
the speed and its feedback to control the motion. Position control is used for the precise
positioning of axes or when a suspended (gravity loaded) axis has to be controlled. Speed
control is used, for example, to control a conveyor belt without position detection or to operate
pumps and fans. In both kinds of actuators, the acceleration, the force or torque and the jerk
can be limited for more realistic behaviour.
Although the parameter that must be controlled in the 3 axes is the position, the drive that
controls the x-axis motor communicates with the PLC using the PROFIdrive standard. In this
case, the position control loop is executed by the PLC and the drive receives speed setpoints.
Therefore, the control of this axis in MCD is modelled as a speed control actuator. The
operation of this drive is explained further in chapter 5.5.1.
In the three axes, the acceleration of the actuators has been limited to have a more realistic
simulation.
5.2. User program
The program that controls the motion of the machine had mostly been developed before the
start of this simulation project. However, the program to control the conveyor motor was not
ready. Hence, a PLC program has been developed and added to the main program by the
author of this report, which can be found in Appendix A. Like this, the whole system can be
simulated to test the machine program once it is ready.
In the physical version of the machine, the position setpoints of each axis are sent to the PLC
by a PC that processes the images of the 3D printed parts captured by the camera and
calculates the movement sequences. To establish the communication between the PC and
PLC, some read and write variables have been defined:
Write variables
- Coordinate X (float): indicates the X-axis setpoint in mm.

- Coordinate Y (float): indicates the Y-axis setpoint in mm.
- Coordinate Z (float): indicates the Z-axis setpoint in mm.
- Counter (unsigned int): verifies that the PC command has been correctly executed by
the PLC.
Read Variables
- Status (unsigned int): indicates the different states of the system (e.g. ready for new
command or homing pending)
- Counter (unsigned int): indicates the completion of an action by the PLC by matching
the value of the Counter variable from the PC.
As explained in chapter 4.1, the camera and the PC are not simulated in this project. Hence,
the position setpoints will be introduced manually through a watch table in TIA Portal shown in
Figure 33. Watch tables allow the user to monitor and modify variables of the program during
its execution.
Figure 33: Watch table in TIA Portal with the variables involved in the PC-PLC communication.
In TIA Portal, apart from programming the PLC, the connection between the drives and the
PLC is defined. The 3 drives are assigned to the PLC network as shown in Figure 34. By doing
so, the input and output addresses of each drive are determined:
- Drive X:
o I address: %I124...141
o Q address: %Q126…135
- Drive Y:
o Q address: %Q200...327
- Drive Z:
o Q address: %Q328...455
These addresses allocate the PLC tags, which are the variables that contain the input and
output signals of the PLC. Those tags are the signals read in SIMIT.
Pg. 40 Report
Figure 34: Network configuration in TIA Portal.
Although the PLC code tested in the simulation will be the same as the one used in the physical
machine, some adjustments have to be made in TIA Portal (the software used to program the
PLC) to support the simulation. Firstly, in the properties of the project, the "Support simulation
during block compilation" checkbox must be enabled.
Secondly, the communication of the drives has to be set to isochronous mode to be able to
use the bus synchronous operation mode described in chapter 4.3.1.2. The cycle time of the
bus is set to 1 ms, which is the minimum. The isochronous mode must also be activated in the
SIMATIC Micro-Drive device configuration and in the program block that controls it (MC-Servo)
as shown in Figure 35 and Figure 36. To complete the isochronous mode configuration, the
topology of the connection between the devices must be set, as shown in Figure 37. In this
way, the PLC can calculate the delays in data transmission.
Figure 35: Tab of the SIMATIC Micro-Drive configuration where the isochronous mode is set.
Figure 36: Settings of the operation mode in the MC-Servo program block.
Figure 37: Topology of the connection between the PLC and the drives.
5.3. Controller simulation
The PLC is simulated in PLCSIM Advanced. When the simulation is started in SIMIT, an
instance of a controller is automatically created in PLCSIM Advanced and set to the "Run"
operating state. Then, the user program in TIA Portal has to be compiled and loaded into the
created virtual controller. After that, the virtual PLC is ready to communicate and synchronise
with SIMIT and MCD.
5.4. Data exchange
SIMIT is used to synchronize the mechatronics model in MCD with the virtual controller in
PLCSIM Advanced. SIMIT automatically detects the input and output signals coming from both
programs. Then, the user can transform and adapt them so that they are understood by the
two software. To enable this communication, two couplings must be configured.
The first coupling is with PLCSIM Advanced. When creating it, it has to be configured as bus
synchronous. The SIMIT project operating mode is also set as bus synchronous. The time
slice used as the simulation cycle time must match the cycle time configured previously in TIA
Portal (1 ms), as shown in Figure 38. In this case, the time slice 2 is used.
Pg. 42 Report
Figure 38: Operating mode and cycle time configuration in the SIMIT project.
The second coupling is with the MCD project. The units of the MCD signals must be set as
they are used in the drive simulation and the PLC, as shown in Figure 39. The MCD coupling
is also set as bus synchronous and the time slice is set to 2.
Figure 39: Signal units in the MCD coupling.
5.5. Drivers behaviour simulation
SIMIT emulates the behaviour of devices like sensors or drives for motors, valves, and pumps.
The models are built using standard libraries or user-made components. The model of each
device is defined using a chart. In this project, three charts have been created to model the
logic of the motor drives.
5.5.1. Drive X
The drive used to control the motor of the conveyor belt is the SIMATIC Micro-Drive PDC from
Siemens.
Siemens drives use the standard PROFIdrive for their communication. PROFIdrive is a
standard for drives that are connected to the controller via PROFInet (explained in chapter
4.2.4). It defines how a drive behaves on the field bus. PROFIdrive devices are activated by
control words (STW) and provide feedback in status words (ZSW) [10]. All the PROFIdrive
drives present a common functionality, shown in Figure 40:
- A state machine
- A ramp generator to transfer the speed setpoint received from the controller
- Adapted feedback of the actual speed and the actual position
Figure 40: Basic functionality of PROFIdrive drives.
The state machine defines the transitions from one status to another. The ramp generator
converts a jump in the speed setpoint into linear ramp increases with an adjustable gradient
[10]. Depending on the telegram used for communication, PROFIdrive devices can also
include a position control or a torque limiting. The telegram used in this project is the telegram
3, which includes a position control. It uses 5 words to send information from the PLC to the
Pg. 44 Report
drive and 9 words to transmit data from the drive to the PLC, which are shown in Figure 41
and explained in Table 2.
Figure 41: Structure of the PROFIdrive telegram 3 [10].
TEL_3 Nominal TEL_3 Actual

STW1 Control word 1 ZSW1 Status word 1
NSOLL_B Standardized speed setpoint NIST_B Standardized actual speed
STW 2 Control word 2 ZSW 2 Status word 2
G1_STW Encoder control word G1_ZSW Encoder status word
G1_XIST 1 Encoder cyclical value
G1_XIST 2 Encoder absolute value
Table 2: PROFIdrive telegram 3 words meaning.
PROFIdrive drives are simulated in SIMIT using standard blocks. Figure 43 shows the model
of the x-axis drive in this project. The blocks used are:
1. PROFIdrive2 component: it simulates the state machine and the communication

according to the PROFIdrive standard. It also sends the speed setpoint to MCD and
reads the actual speed coming back. The reference and maximum speed of this
component must be coherent with the motor characteristics and the configuration of
the motor in TIA Portal. The ramp time is set to 0 because, in this case, the PLC is
managing the acceleration and deceleration ramps. Figure 42 shows the PROFIdrive2
block parameters.
Figure 42: PROFIdrive2 block parameters.
2. Sensor component: it is the connection to an encoder. The resolution parameters must

be set according to the encoder used.
3. SensorProcessRotatory component: it simulates a rotary encoder, either absolute or

incremental. It converts a signal in degrees coming from MCD to an encoder signal. In
this case, the encoder is incremental, and it has 2048 steps per revolution. These are
the same parameters set in TIA Portal.
Figure 43: Model of the x-axis drive in SIMIT.
These PROFIdrive blocks exchange information with MCD and the virtual controller using the
words in Table 2 plus NIst, NSoll and Gx_XIst. NIst is the actual speed in min-1, NSoll is the
speed setpoint in min-1 and Gx_XIst is the motor position feedback in degrees.
As the encoder used is incremental, the axis has to be referenced at the beginning of the
operation. To do so, the binary signal of the collision sensor in MCD is sent to the
corresponding input of the PLC, as seen at the bottom of Figure 43.
5.5.2. Drive Y and Z
The drive used to control the Y and Z motors is the Lexium MDrive integrated motor by
Schneider Electric. Although it is connected to the controller via PROFInet, it does not comply
with the PROFIdrive standard. Thus, a specific model has to be developed to simulate its
behaviour. As there is no documentation available explaining the logic of this drive (state
machine, control and status words, etc.), a simplified model has been built by a reverse
engineering process. The outcome of understanding the signals and the necessary
Pg. 46 Report
transformations to get a correct functioning of the simulation can be seen in Figure 44 for the
y-axis drive, and in Figure 45 for the z-axis drive.
Figure 44: Model of the y-axis drive in SIMIT.

Figure 45: Model of the z-axis drive in SIMIT.
The PLC tags used in the data exchange between the Lexium MDrive and the PLC have a
specific data structure, called “TLexium_MDrv_Drv2Plc” for the PLC input signals and
“TLexium_MDrv_Plc2Drv2” for the PLC output signals, as seen in Figure 46. Each of these
tags is a folder with a list of other tags that are used for the operation of the motor and the
signal exchange with the PLC.
Figure 46: PLC tags for data exchange with the Lexium MDrive.
Pg. 48 Report
The tag DE carries a byte signal that enables the drive. If the lower bit of this signal is set to
one, the position and speed setpoints sent by the PLC are transferred to the MCD model. The
position setpoint is set by the tag MA (DInt). It is an absolute position. The speed setpoint is
set by the tag VM (DWord). The maximum acceleration and deceleration in the MCD
movements are set by the tags A (DWord) and D (DWord).
The position, speed, acceleration and deceleration signals sent by the PLC are integer values
in the units understood by the real drive and motor (micro-steps). Those signals must be
converted to analogue values and change the units to the ones understood by the mechatronic
model in MCD. Each motor revolution equals to 51200 micro-steps.
The drive model also applies the transmission ratio. The linear travel per motor revolution is
70 mm/rev in the y-axis and 72,25 mm/rev in the z-axis.
The position feedback from the mechatronic model is transformed and linked to the PLC signal
P (DInt), simulating the incremental encoder.
Finally, the tag MV (Byte) monitors the movement of the motor. In the real drive, it is set to 1 if
the motor is moving to a position or 0 if it is stopped. When the PLC sees a change of this tag
from 1 to 0, it sets the action that was being performed by the drive as completed. To simulate
this behaviour in SIMIT, the signal MV is set to 1 when one of the following conditions is true:
- The speed of the motor in the mechatronic model is out of the range of -0,1 and 0,1
mm/s.
- The PLC program has requested a referencing of the axis and the home sensor is
activated. When a homing operation is needed, the PLC triggers the tag
Toggle_reg_22.
As the simulation will start with the three axes at the home position, the movements of the
homing operation are not simulated.
6. Validation of the simulation
6.1. Operation of the simulation
Once the simulation is ready, it can be launched from SIMIT. MCD starts automatically and
shows the machine. At the same time, a virtual controller is created and started in PLCSIM
Advanced. Then, the user program has to be compiled and loaded to the virtual PLC. After
that, the simulation can be operated using the watch table in TIA Portal explained in
chapter 5.2, which contains the variables for the communication between the PLC and the
PC in the real machine.
The control code for the x-axis has been added during this project and with the sole
purpose to test the simulation. As this portion of code will not be part of the final machine
control program, it has been decided not to dedicate further efforts to integrate it with the
PC-PLC communication. Therefore, while the PC-PLC communication variables trigger
the y and z movements and provide feedback on their completion, they do not reflect the
x-axis motion status.
At the start of the simulation, the value of the ComPLCPC.Status variable is 3. The user
program requires referencing the z and y axes, in this order, before accepting any position
setpoint. The homing of the z-axis is triggered by setting the ComPCPLC.Counter to 1.
Once the z-axis is homed, the PLC sets ComPLCPC.Counter to 1 and the
ComPLCPC.Status to 4, as in Figure 47.
Figure 47: Values of the communication variables after referencing the z-axis.
To reference the y-axis, the user has to set the ComPCPLC.Counter to 2. After the home
sensor is triggered and the actual position is set to 0, the communication variables have
the values shown in Figure 48.
Figure 48: Values of the communication variables after referencing the y-axis.
Pg. 50 Report
The x-axis is enabled and homed at the start of the simulation. However, it can not start moving
until the other two axes are homed.
Once the 3 axes are referenced, the user can start introducing position setpoints in the
ComPCtoPLC.CoordinateX, ComPCtoPLC.CoordinateY and ComPCtoPLC.CoordinateZ
variables and observe the machine motion in MCD. Each time a new position command is
introduced, the user has to change the value of ComPCPLC.Counter to an integer bigger than
2. After the y and z axes reach their target positions, the ComPLCPC.Counter matches the
ComPCPLC.Counter.
[17] provides a video of the operation of the simulation and the resulting visualization of the
motion of the machine. It can be observed that the 3 axes respond to the position commands
as expected.
The default values for the maximum linear speed, acceleration and deceleration used in the
machine program are shown in Table 3. The y and z acceleration and deceleration values are
equivalent to 100000 micro-steps/s2.
Axis Speed [mm/s] Acceleration [mm/s2] Deceleration [mm/s2]

X 30 600 600
Y 10 137 137
Z 10 141 141
Table 3: Default values for linear speed, acceleration and deceleration.
For a deeper analysis of the accuracy of the simulation, the position, speed and acceleration
values versus time of each axis are exposed and discussed in the following chapter.
6.2. X-axis motion
To test x-axis motion, four tests have been conducted with different values for the position,
speed, acceleration and deceleration. In all the tests, the movement starts from the 0 mm
position. Then, the position command is introduced. Once it is reached, the position setpoint
is again changed to 0. The characteristics of each movement can be seen in Table 4.
Target Maximum Maximum Maximum

Test position speed acceleration deceleration Comment
[mm] [mm/s] [mm/s2] [mm/s2]
1 750 30 600 600 Default values
2 5 30 600 600 Close position
3 750 55 600 600 High speed
4 750 30 60 30 Low acceleration and deceleration
Table 4: Characteristics of the x-axis motion tests.
During the performance of the tests, it has been detected that, with a simulation cycle time of
4 ms, the x-axis model has an unexpected behaviour when it is set to speeds higher than 1875
min-1. This error has been communicated to Siemens technical support. Their suggestion is to
lower the simulation cycle time to 1 ms, which is the solution adopted in this project. The cause
of the faulty behaviour is related to the software and it is currently under investigation by
Siemens.
Figure 49 shows the motion graph obtained using the default settings of the x-axis motor. The
30 mm/s speed corresponds approximately to the motor nominal speed of 3450 min-1. In Figure
49 (c) it seems like the maximum acceleration values (600 mm/s2) are not reached. However,
this is due to the discrete data collection. With a sample time of 0,03 s, the moments where
the maximum acceleration is reached are not recorded.
(a)
(b)
(c)
Figure 49: X-axis motion with a target position of 750 mm, a maximum speed of 30 mm/s and
maximum acceleration and deceleration of 600 mm/s2. (a) Position, (b) Speed and (c) Acceleration
versus time.
Pg. 52 Report
Figure 50 shows the motion profile of the second test, where the target position is 5 mm. As
seen in Figure 50, the target position is reached in less than 0,05 s. This does not cause any
instability in the simulation.
(a)
(b)
(c)
Figure 50: X-axis motion with a target position of 5 mm, a maximum speed of 30 mm/s and maximum
acceleration and deceleration of 600 mm/s2. (a) Position, (b) Speed and (c) Acceleration versus time.
The speed used in test 3 corresponds to a motor speed of approximately 6500 min-1. This
speed can not be reached by this motor. However, the test was performed to check if the
model presented unexpected behaviours at high speeds. As seen in Figure 51, the motion
profile obtained is correct.
(a)
(b)
(c)
Figure 51: X-axis motion with a target position of 750 mm, a maximum speed of 55 mm/s and
versus time.
Figure 52 shows how the simulation responds to getting different values for acceleration and
deceleration. The speed profile obtained in the four tests is trapezoidal, as the target position
is far enough to give time to reach the maximum speed set in each case.
Pg. 54 Report
(a)
(b)
(c)
Figure 52: X-axis motion with a target position of 750 mm, a maximum speed of 30 mm/s, a maximum
acceleration of 60 mm/s2 and a maximum deceleration of 30 mm/s2. (a) Position, (b) Speed and (c)
Acceleration versus time.
6.3. Y-axis motion
To test y-axis motion, four tests have been conducted with different values for the position,
speed, acceleration and deceleration. The movement sequence is the same as for the x-axis.
The characteristics of each movement can be seen in Table 5.

[mm] [mm/s] [mm/s ] 2
[mm/s ] 2
1 280 10 137 137 Default values

2 2 10 137 137 Low position
3 280 1075 137 137 High speed
4 280 10 14 7 Low acceleration and deceleration
Table 5: Characteristics of the y-axis motion tests.

Figure 53 shows the motion graph obtained using the default settings of the y-axis motor. With
an acceleration of 137 mm/s2, a speed of 10 mm/s is reached in 0,07 s. Since the data
recording sample time is 0,03 s, the speed change is only represented by two points where
the acceleration is not 0, as shown in Figure 53 (c).
(a)
(b)
(c)
Figure 53: Y-axis motion with a target position of 280 mm, a maximum speed of 10 mm/s and
versus time.
In Figure 54 (c), the acceleration phase is represented by two or three points like in the previous
case. They are more visible because, as the movement is shorter, the amount of data is lower,
and the points are more spaced.
Pg. 56 Report
(a)
(b)
(c)
Figure 54: Y-axis motion with a target position of 2 mm, a maximum speed of 10 mm/s and maximum
With a higher speed or a lower acceleration and deceleration, the motion profile is smother, as
in Figure 55 and Figure 56. The speed of 1075 mm/s in test 3 is approximately 786000 micro-
steps/s or 921,1 min-1, which is the default maximum speed value for the Lexium MDrive motor
drive. As seen in Figure 55, the actuator responds to this parameter with a triangular speed
profile that has its vertex at the speed of 194,71 mm/s. Therefore, the maximum speed is not
reached in this movement, although it could be reached if the target position was further.
However, 280 mm is the limit position of the y-axis.
(a)
(b)
(c)
Figure 55: Y-axis motion with a target position of 280 mm, a maximum speed of 1095 mm/s and
versus time.
Figure 56 shows how the simulation responds to getting different values for acceleration and
deceleration. Figure 56 (c) shows that the model processes them correctly. In test 1, 2 and 4,
the speed profile is trapezoidal, as the target position is far enough to give time to reach the
maximum speed of 10 mm/s.
Pg. 58 Report
(a)
(b)
(c)
Figure 56: Y-axis motion with a target position of 280 mm, a maximum speed of 10 mm/s, a maximum
6.4. Z-axis motion
To test z-axis motion, four tests have been conducted with different values for the position,
speed, acceleration and deceleration. The movement sequence is the same as for x and y
axes. The characteristics of each movement can be seen in Table 6.

[mm] [mm/s] [mm/s2] [mm/s2]
1 -240 10 141 141 Default values
2 -2 10 141 141 Close position
3 -240 1009 141 141 High speed
4 -240 10 14 7 Low acceleration and deceleration
Table 6: Characteristics of the z-axis motion tests.

Figure 57, Figure 58, Figure 59 and Figure 60 show the resulting motion graphs of the tests.
The same observations done in the y-axis are applicable for the z-axis. Note that the position
of the z-axis is negative.
(a)
(b)
(c)
Figure 57: Z-axis motion with a target position of 240 mm, a maximum speed of 10 mm/s and
versus time.
Pg. 60 Report
(a)
(b)
(c)
Figure 58: Z-axis motion with a target position of 2 mm, a maximum speed of 10 mm/s and maximum
(a)
(b)
(c)
Figure 59: Z-axis motion with a target position of 240 mm, a maximum speed of 1009 mm/s and
versus time.
Pg. 62 Report
(a)
(b)
(c)
Figure 60: Z-axis motion with a target position of 240 mm, a maximum speed of 10 mm/s, a maximum
6.5. Discussion
After analysing the results, it can be concluded that the motion graphs are coherent and match
with the expected physical behaviour of the machine.
At this point, the simulation can be used to evaluate the performance of different speed and
acceleration values and optimize the pick and place algorithms. The default speed and
acceleration values used in the z and y axes are quite low because they are selected to reduce
the risks of unexpected collisions when doing the tests in the real machine. However, they are
not optimum to pick and place a big amount of parts, as the operation cycle time is very high.
As pick and place applications do not require an area with constant speed, like machining
operations, they usually use triangular speed profiles. Triangular speed profiles also provide
the fastest movement between two points. They are achieved when using high maximum
speeds, like in tests number 3 of the y and z axes. Therefore, the virtual model can be used to
validate high-speed movement sequences and optimize cycle times before running them in
the physical machine. This can be done remotely, without requiring expensive and scarce
hardware to run the experiments. By doing so, the real commissioning phase can be reached
with higher quality and more mature code. Likewise, the commissioning time can be shortened
as most of the debugging will be already completed.
The benefits pointed out for this project apply to many other industry cases where movement
sequences have to be verified and optimized. To name a few, machining, sorting, packaging
or cleaning operations.
Pg. 64 Report
7. Economic analysis
The costs of this project are gathered in Table 7. They are associated with labour, software
licenses and the hardware to run the software.
Quantity Yearly cost Hourly Total

Concept
[h] [€/year] cost [€/h] cost [€]
Junior engineer 360 28000 13,83 4980,24
NX Mechatronics Concept Designer
360 25572 12,63 4548,38
license
SIMIT 10.1 M license 360 6125 3,03 1089,43
S7-PLCSIM Advanced V2.0 SP1 + TIA
360 5000 2,47 889,33
Portal V15.1 license
Microsoft Office 365 license 360 130 0,06 23,12
Personal computer 360 667 0,33 118,58
Total 11649,07
Table 7: Costs of the project.
The length of this project is calculated considering that it is comparable to the 12 ECTS of the
Master Final Thesis. Assuming that a 1 ECTS is approximately 30 h working hours, this project
is equivalent to 360 h. The hourly cost of all the concepts is calculated assuming 8-hours
workdays in a calendar with 253 workdays per year. The total cost comes from multiplying the
hourly cost per the number of hours of this project.
The labour costs are calculated assuming an annual gross salary for a junior industrial
engineer of 28000 €.
The software licenses used to conduct the simulation project are:
- NX Mechatronics Concept Designer

- SIMIT 10.1 M
- S7-PLCSIM Advanced V2.0 SP1
- TIA Portal V15.1
The yearly prices in Table 7 correspond to the cost of a floating license with a yearly
subscription. Those prices are an approximation.
To write this report, multiple software of the Microsoft Office 365 professional package has
been used. The price in Table 7 corresponds to the cost of a yearly subscription for one user.
Finally, this thesis has been conducted and written in a 2000 € personal computer. Assuming
a computer life of 3 years, the cost per year is 667 €.
The resulting total cost of this project is 11649,07 €. This cost can be understood as the
necessary initial investment to acquire the tools and the knowledge to develop a system with
a model-based methodology. It costs 30,83% more than the cost of building a prototype,
estimated in Table 8.
Concept Quantity [h] Hourly cost [€/h] Cost [€]

Conveyor - - 714,00
Cartesian robot - - 3950,00
Tray - - 1000,00
Structure - - 840,00
Mechanic technician 80 15 1200,00
Electrical technician 80 15 1200,00
Total 8904,00
Table 8: Costs of building the prototype.
When it comes to the time, the number of hours devoted to this project are 360 h, which are
equivalent to 9 weeks working 40 hours per week. However, it is estimated that after gaining
the knowledge to build the first model, the time required to create a new one would be between
2 and 4 weeks. Therefore, the cost of the project would be also divided by 0,44 in the worst
case, resulting in 3957,33 €.
The assembly of the prototypes usually takes 3 weeks, while the lead time between the order
and the reception of the manufactured components can vary between 2 and 8 weeks. During
the manufacturing, electric engineers prepare the wiring diagrams and the I/O assignment,
which takes 3 weeks. After that, the simulation can be developed. Therefore, there is more
than enough time and human resources to parallelize the development of the model while the
machine components are being manufactured and assembled. If the commissioning is also
started before the assembly is finished, which typically takes 3 or 4 more weeks, the total
project length can be reduced. Figure 61 shows a possible Gantt chart of the project
introducing virtual commissioning.
Pg. 66 Report
Figure 61: Gantt chart of the machine-building process using virtual commissioning.
On the other hand, if hardware errors are detected once the machine is built, reworks or new
parts might be needed, and thus an additional cost and delay in the machine start-up.
Validating the machine concepts in a virtual prototype before the machine is approved for
manufacturing can help reduce these unexpected expenses and time.
All in all, despite the initial investment required, adopting a model-based development
approach shortens the project timeline and potentially brings cost-savings in the long term.
8. Environmental impact
The environmental impact associated with this project comes from the energy consumption of
the personal computer and the CO2 emissions generated to produce this energy.
Figure 62 and Figure 63 show the evolution of the electric energy generation and the CO2
emissions in the Spanish Peninsula from the 11th of April 2020 at 3 am till the 12th of April 2020
at 3 am [16].
7000
Wind
6000 Nuclear
Electric energy generation [MW]
Fuel/gas
5000
Coal
4000 Combined
cycle
3000 Hydro
Int.
2000 exchanges
Balear link
1000 Solar PV
0 Solar thermal
Thermal
-1000 renewable
Cogeneration
03:00 Apr.11 08:00 Apr.11 13:00 Apr.11 18:00 Apr.11 23:00 Apr.11 and waste
Date and Time
Figure 62: Evolution of the electric energy generation by central type along one day in the Spanish
Peninsula.
1600
Wind
1400 Nuclear
Fuel/gas
CO2 emissions [t CO2 eq/h]
1200
Coal
1000 Combined cycle
800 Hydro
Int. exchanges
600
Balear link
400 Solar PV
Solar thermal
200
Thermal renewable
0 Cogeneration and
Apr.11 03:00 Apr.11 08:00 Apr.11 13:00 Apr.11 18:00 Apr.11 23:00 waste
Date and time
Figure 63: Evolution of the CO2 emissions by central type along one day in the Spanish Peninsula.
Pg. 68 Report
Figure 64 shows the resulting CO2 emissions caused by the generation of electric energy along
one day in the Spanish Peninsula. As it varies a lot, the average of 35,62 gCO2 eq/MJ is taken
for the calculations.
50,00
CO2 emissions by generation of electric
45,00
40,00
energy [g CO2 eq/MJ]
35,00
30,00
25,00
20,00
15,00
10,00
5,00
0,00
Apr.11 03:00 Apr.11 08:00 Apr.11 13:00 Apr.11 18:00 Apr.11 23:00
Date and time
Figure 64: Evolution of the CO2 emissions by the generation of electric energy along one day in the
Spanish Peninsula.
As the power of the personal computer is 150W and it is used for 360h, the resulting CO2
emissions are shown in Table 9.
Power [kW] Usage time [h] Energy [kWh] CO2 Emissions [g CO2 eq]
0,15 360 54 6924,96
Table 9: CO2 emissions associated with the use of the personal computer.
Conclusion
The results obtained from the validation of the simulation proof that the model provides a good
representation of the physical machine. The simulated physics, behaviour of the motor drives
and data exchange perform as expected when the control program is executed in the virtual
controller.
The use case presented in this project shows the advantages of using a model-based
development approach. Firstly, the machine can be commissioned before the real hardware
is assembled. This enables parallelization of tasks. For example, automation engineers can
develop the control programs and test them while the components are being manufactured
and the machine is being built. As they arrive at the real commissioning phase with an already
validated system, it becomes be much faster to get the machine up and running to do further
testing or deliver it to customers.
Testing in a virtual model also enables early error identification and reduces the risk of
damaging real equipment while the machine operation is validated. In the application of this
project, for instance, multiple grippers could be tested without the danger of breaking 3D printer
parts. Likewise, it can be used to validate high-speed movement sequences, which could
cause unexpected collisions if they are tested in the real hardware first.
By having a digital twin of the system, the control software can be developed from multiple
locations. This is particularly useful in a multinational company like HP, where most of the
projects are co-developed by international teams. It also decreases the investment needed in
hardware for testing. Likewise, it provides the opportunity to outsource some of the
development to third parties. In the use case of this project, the virtual model can be used to
develop and validate pick and place algorithms for customers without having access to their
facilities or 3D printed parts. It must be pointed out that having multiple users working with the
simulation requires having a version tracking system to prevent developers from working with
faulty or outdated models.
Using virtual system models can also help in the early stages of development. Machine
concepts and operation sequences can be validated in tools like NX Mechatronics Concept
Designer without having to develop the electronics or the software. If the concept does not
work, the equipment is not built. However, this does not remove the need to build prototypes
to identify performance characteristics that can not be simulated in the virtual model.
These tools also make it easier to agree upon the functionality of the system between various
disciplines. The integration issues are reduced as does the documentation of the project, as
the simulation serves as the documentation.
Pg. 70 Report
When it comes to customers, the virtual model can be a tool to demonstrate the proposed
system to the client. Then, the client can suggest changes and the machine manufacturer can
implement them at a minor cost before the system is built. In this way, the system builder also
makes sure the final product is aligned with customers’ requirements.
Finally, model-based development opens the door to test automation and automatic
generation of code, which would make the development process even more efficient.
Despite the multiple advantages of model-based development and virtual commissioning,

some stoppers still prevent companies from adopting this methodology. In the beginning, it
needs an investment in the simulation tools and the know-how to create the models, which is
something not paid by the customer. Furthermore, the amount of time and effort needed to
develop and fine-tune the first models is significant. In this project, a big part of the time was
invested in understanding and modelling the Lexium MDrive, as no standard simulation block
was available. However, once the model of a component is created, it can be reused as many
times as desired, reducing the time required to perform future simulations.
This project ends with the delivery of a simulation ready to understand the position, speed and
acceleration values introduced by the user and provide a correct visualization of the machine
operation. With this, the scope of this work is covered. However, to simulate the complete
system described in chapter 4, the following future work would be needed:
- Establish the communication between the virtual controller and the PC that calculates
the pick and place algorithms. This would allow obtaining the target coordinates
automatically instead of introducing them manually.
- Add the packaging box and the 3D printer parts detected by the camera in the physics
simulation. This would enable detecting collisions between these objects and the
machine.
- Develop the program to control the motion of the x-axis, the gripper to pick and place
the parts and the machine door sensors that prevent access to the machine when it is
running. This would be done by automation engineers.
- Model the physics of the gripper and the behaviour of the driver that controls it.
- Model the behaviour of the machine door sensors.
Once this work is completed, automation engineers will be able to test in a virtual model the
pick and place movement sequences for each part, as well as the operation of the whole
machine.
Acknowledgements
I would like to thank Pablo Murciego for introducing me to model-based development
processes and mindset and guiding me through the election of the topic of this project. Also,
for supporting the development of this work by providing all the licenses and resources needed.
I would also like to thank Lluïsa Jordi for her support and advice on the format and language
of this report and the academical suitability of this project.
I would also like to acknowledge Álvaro Hernández for his technical support regarding the
Siemens simulation toolset.
Likewise, I would like to thank Blanca de Prada, Jose Juan Hidalgo and Carles Centeno for
helping me understand the control program and operation of the machine simulated in this
project.
Finally, I would like to thank my family and friends for their continuous and unconditional
support.
Pg. 72 Report
Bibliography
Citations
[1] A. FERNÁNDEZ, M. A. EGUÍA AND L. E. ECHEVERRÍA, Virtual commissioning of a

robotic cell: an educational case study, 2019 24th IEEE International Conference on
Emerging Technologies and Factory Automation (ETFA), Zaragoza, Spain, 2019, pp.
820-825. [https://ieeexplore-ieee-
org.recursos.biblioteca.upc.edu/document/8869373/references#references]
[2] H. VERMAAK AND J. NIEMANN, "Virtual commissioning: A tool to ensure effective

system integration," 2017 IEEE International Workshop of Electronics, Control,
Measurement, Signals and their Application to Mechatronics (ECMSM), Donostia-San
Sebastian, 2017, pp. 1-6. [https://ieeexplore-ieee-
org.recursos.biblioteca.upc.edu/document/7945899/references#references]
[3] REINHART, G., WÜNSCH, G. Economic application of virtual commissioning to

mechatronic production systems, Prod. Eng. Res. Devel. 1, 371–379, 2007. [https://doi-
org.recursos.biblioteca.upc.edu/10.1007/s11740-007-0066-0]
[4] SHAHIM, N., MØLLER, C. Economic justification of Virtual Commissioning in

automation industry, 2016 Winter Simulation Conference (WSC), Washington, DC,
2016, pp. 2430-2441. [https://ieeexplore-ieee-
org.recursos.biblioteca.upc.edu/document/7822282]
[5] BORSTELL, H., REGGELIN, T. Towards Virtual Commissioning of Image-based

Information Systems for State Detection in Logistics. 9th IFAC Conference Manufacturing
Modelling, Management and Control MIM 2019, Berlin, 2019.
[https://www.researchgate.net/publication/336369277_Towards_Virtual_Commissionin
g_of_Image-based_Information_Systems_for_State_Detection_in_Logistics#pf3]
[6] ARC ADVISORY GROUP. Siemens PLM Software’s Advanced Machine Engineering
Streamlines Machine Development, ARC White Paper, 2015.
[https://www.plm.automation.siemens.com/media/global/en/AME_-
_ARC_White_Paper%20-%20Advanced_Machine_Engineering_Streamlines_Machine
_Development_tcm27-53754.pdf, 29th of February 2020]
[7] ITQ, ITQ Corporate Brochure, 2018. [https://www.itq.de/wp-

content/uploads/2019/06/ITQ_corporate_brochure.pdf, 29 of February 2020]
th
[8] SIEMENS AG, Address critical design, engineering and manufacturing challenges with
Advanced Machine Engineering, Webinar.
[https://www.plm.automation.siemens.com/global/en/webinar/advanced-machine-
engineering/25892, 29th of February 2020]
[9] SIEMENS AG, SIMATIC Machine Simulator V2.0. Getting started.

[https://support.industry.siemens.com/cs/attachments/109758943/s71500_simatic_mac
hine_simulator_getting_started_v2.0_en-US.pdf, 21st of March 2020]
[10] SIEMENS AG, SIMIT Simulation Platform (V10.1). Operation Manual.

[https://support.industry.siemens.com/cs/attachments/109759317/SIMIT_enUS_en-
US.pdf, 21st of March 2020]
[11] SIEMENS AG, Mechatronic Concept Designer.

[https://www.plm.automation.siemens.com/global/en/products/mechanical-
design/mechatronic-concept-design.html, 21st of March 2020]
[12] HP INC., HP Metal Jet technology, Technical white paper.

[https://h20195.www2.hp.com/v2/GetDocument.aspx?docname=4AA7-3333ENW, 8th of
March 2020]
[13] LINK3D, Link3D Additive Manufacturing Technologies Overview.

[https://solution.link3d.co/3d-printing-resources.html, 7 of March 2020]
th
[14] PLC DEV, How PLCs Work. [http://www.plcdev.com/how_plcs_work, 7th of March 2020]
[15] SIEMENS PRODUCT LIFECYCLE MANAGEMENT SOFTWARE INC., Siemens

Documentation: Mechatronics Concept Designer, 2019.
[https://docs.plm.automation.siemens.com/tdoc/nx/1872/nx_help/#uid:index_mechatron
ics, 10th of April 2020]
[16] RED ELECTRICA DE ESPAÑA, Spanish Peninsula – Electricity demand tracking in real-
time. [https://demanda.ree.es/visiona/peninsula/demanda/total/2020-04-10, 13th of April
2020]
[17] CARBONELL, B. Simulation with MCD, SIMIT and TIA Portal.

[https://youtu.be/Ulq8Oe2rYSI, 18 of April 2020]
th
Complementary bibliography
[18] C. HENKE, J. MICHAEL, C. LANKEIT AND A. TRÄCHTLER, A holistic approach for

virtual commissioning of intelligent systems: Model-based systems engineering for the
Pg. 74 Report
development of a turn-milling center, 2017 Annual IEEE International Systems

Conference (SysCon), Montreal, QC, 2017, pp. 1-6. [https://ieeexplore-ieee-
org.recursos.biblioteca.upc.edu/document/7934735]
[19] GUERRERO, L., LÓPEZ, V., MEJÑIA, J., Virtual Commissioning with Process
Simulation (Tecnomatix), Computer-Aided Design and Applications, 11:sup1, S11-S19,
2014. [https://www.tandfonline.com/doi/full/10.1080/16864360.2014.914400]
[20] DEFENSE ACQUISITION UNIVERSITY PRESS, Systems Engineering Fundamentals,

2001. [https://ocw.mit.edu/courses/aeronautics-and-astronautics/16-885j-aircraft-
systems-engineering-fall-2005/readings/sefguide_01_01.pdf, 29th of February 2020]
[21] SIEMENS AG, SIMATIC S7-1500T - Virtual commissioning for kinematics in NX MCD
with Software in the Loop, January 2019.
[https://support.industry.siemens.com/cs/document/109760078/simatic-s7-1500t-virtual-
commissioning-for-kinematics-in-nx-mcd-with-software-in-the-loop-?dti=0&lc=en-US,
21st of March 2020]

Report TFM Bertacarbonell

Uploaded by

Copyright:

Available Formats

Report TFM Bertacarbonell

Uploaded by

Document Information

Original Title

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

Report TFM Bertacarbonell

Uploaded by

Copyright:

Available Formats

Master Final Thesis

Master’s Degree in Industrial Engineering

Usage of a mechatronics software for efficient machine

Author: Berta Carbonell i Vilaplana

Escola Tècnica Superior

The virtual model is validated by a 3D visualization of the machine motion as well as an

3. STATE OF THE ART ______________________________________ 13

4. SYSTEM DESCRIPTION ___________________________________ 20

4.3.1.1. Time slices ........................................................................................... 29

4.3.1.2. Operation modes .................................................................................. 29

5. DEVELOPMENT OF THE DIGITAL TWIN ______________________ 33

5.1.1.1. Rigid bodies ......................................................................................... 33

5.1.1.2. Collision bodies .................................................................................... 35

5.1.3. Couplers ..........................................................................................................36

6. VALIDATION OF THE SIMULATION __________________________ 49

7. ECONOMIC ANALYSIS ____________________________________ 64

PLC: Programmable Logic Controller

VC: Virtual Commissioning

HMI: Human-Machine Interface

PC: Personal Computer

TO: Technology Object

it in the physical hardware.

2.2.1. Usage of programmable logic controllers

model created with some of the PLC simulation tools.

2.2.2. HP Metal Jet technology

HP Metal Jet is an additive manufacturing technology for mass-production. Thus, it is important

Figure 1: Additive Manufacturing Technologies [13].

Figure 2: Schematic of HP Metal Jet additive manufacturing process [12].

3. State of the art

3.1. Machine development process

To reduce project timelines, machine manufacturers are introducing innovations in internal

3.1.1. Traditional sequential engineering

Figure 3: Traditional machine development process [8].

Figure 4: Traditional, sequential engineering process [7].

Figure 5: Evolution of software content of machines [7].

Some of these drawbacks are:

- Requirements mismatch: since customers' requirements are only revised at the

3.1.2. Systems engineering and concurrent design

Figure 6: Today’s machine development process [8].

Figure 7: Engineered systems development process [7].

3.2. Virtual Commissioning

Virtual commissioning consists of using simulation tools to emulate a physical system

Figure 8: Virtual commissioning of control systems [5].

Virtual commissioning can be performed in two configurations: hardware-in-the-loop and

3.2.1. Hardware in the loop

Hardware-in-the-loop (HIL) configuration requires having a virtual model of the process or

Figure 9: Hardware-in-the-loop (HIL) configuration.

3.2.2. Software in the loop

In a software-in-the-loop (SIL) configuration, the model of the system is controlled by a virtual

Figure 10: Software-in-the-loop (SIL) configuration.

3.2.3. Virtual commissioning software

Figure 11 provides an overview of the simulation levels of Siemens virtual commissioning

Figure 11: Simulation levels of Siemens virtual commissioning software [8].

Parts tray Packaging box

Figure 13: Top view of a schematic representation of the system architecture.

Figure 14: Front view of a schematic representation of the system architecture.

4.2. Hardware architecture

4.2.1. Cartesian robot

Figure 15: 2-axis cartesian robot at home position