You are on page 1of 12

ScienceDirect

Available online at www.sciencedirect.com

ProcediaScienceDirect
Manufacturing00 (2018) 000–000
Available www.elsevier.com/locate/procedia
Availableonline
onlineatatwww.sciencedirect.com
www.sciencedirect.com
Procedia Manufacturing00 (2018) 000–000
www.elsevier.com/locate/procedia
ScienceDirect
ScienceDirect 
Procedia Manufacturing 26 (2018) 95–106
Procedia Manufacturing 00 (2017) 000–000
46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
www.elsevier.com/locate/procedia

46th SME North American Manufacturing Research Conference, NAMRC 46, Texas, USA
Modelling, Simulation and Control of Incremental Sheet Metal
Modelling,Forming
SimulationProcess using CNC
and Control Machine Tool
of Incremental Sheet Metal
Ketul Patel ,V.Kalaichelvi
a
Manufacturing FormingSociety
Engineering b*
Process,R.Karthikeyan
using CNC
International b
,Sriparvathi
Machine
Conference Bhattathiri
Tool2017,
2017, MESIC 28-30 Juneb
2017, Vigo (Pontevedra), Spain
Ketul Patela,V.Kalaichelvib*,R.Karthikeyanb,Sriparvathi Bhattathirib
a University of Michigan, USA b BITS Pilani, Dubai Campus., Dubai.,UAE * Corresponding author. Tel.: +971 55 6301853;

Costing models for capacity optimization in Industry 4.0: Trade-off


E-mail address: kalaichelvi@dubai.bits-pilani.ac.in
a University of Michigan, USA b BITS Pilani, Dubai Campus., Dubai.,UAE * Corresponding author. Tel.: +971 55 6301853;

between used capacity and operational efficiency


E-mail address: kalaichelvi@dubai.bits-pilani.ac.in

Abstract
A. Santanaa, P. Afonsoa,*, A. Zaninb, R. Wernkeb
Abstract
Incremental sheet forming (ISF) is an advanced metal forming technology. Local progressive deformation of the metal sheet is
achieved to get desired shape through ISF. a
It is gaining
University popularity
of Minho, 4800-058in Guimarães,
customizedPortugal
production. However, ISF has lower geometric
Incremental sheet forming (ISF) is an
accuracy compared to the conventional sheet advanced
b
metal forming
Unochapecó, 89809-000
forming technology.
Chapecó,such
methods Local
SC, as progressive
Brazil deformation
bending, deep drawing ofandthestamping.
metal sheet is
The
achieved toaccuracy
geometric get desired
canshape through by
be improved ISF. It is gaining
adopting popularity
suitable control in customized
system strategyproduction.
in machineHowever, ISF this
tool. Hence has paper
lower deals
geometric
with
accuracy
the compared
modelling to the conventional
and simulation of ISF CNCsheet machinemetaltoolforming methods
to address suchThe
this issue. as CNC
bending, deepmachine
milling drawingmodel
and was
stamping. The
developed
geometric
using CADaccuracy can be improved
and kinematics. by adopting
A mechanistic model suitable control by
was developed system strategyseparately
integrating in machine tool. Hence
developed modelsthisofpaper deals with
all mechanical
Abstract
the
and modelling and simulation
electrical components of ISF
of CNC CNC machine
machine tool tocontrol
tool. Cascade address this issue.
strategies The
were CNC and
applied milling machine
compared model
to find thewas
mostdeveloped
accurate
using CAD
one. The bestand kinematics.
control strategyAwas
mechanistic model
also validated forwas
ISFdeveloped
using two by integrating
different separately
symmetric sheet developed
metal CADmodels
models. of all mechanical
Under the concept
and electrical componentsof "Industry 4.0", tool.
of CNC machine production
Cascade processes will be
control strategies werepushed
applied to
andbe increasingly
compared interconnected,
to find the most accurate
information
one. The best control strategy was also validated for ISF using two different symmetric sheet metal CAD models. optimization
based on a real time basis and, necessarily, much more efficient. In this context, capacity
goes beyond
© 2018 the traditional
The Authors. Published aim
by of capacity
Elsevier B.V.maximization, contributing also for organization’s profitability and value.
© 2018 The Authors. Published by Elsevier B.V.
Indeed, lean
Peer-review undermanagement
under responsibility and
responsibilityof ofthecontinuous
thescientific improvement
scientificcommittee
committee approaches suggest capacity optimization instead of
Peer-review ofofthe
NAMRI/SME.
46th SME North American Manufacturing Research Conference.
maximization. The study
© 2018 The Authors. of capacity
Published by Elsevier optimization
B.V. and costing models is an important research topic that deserves
Peer-review
contributionsunder
from
Keywords:Incrementalresponsibility
both Metal
Sheet of the scientific
the practical
Forming; CNC committee
and Machine
theoretical of NAMRI/SME.
Tool;perspectives.
Control Systems;This paper presents
Kinematics; and discusses a mathematical
System Modelling
model for capacity management based on different costing models (ABC and TDABC). A generic model has been
Keywords:Incremental
developed and it wasSheet
usedMetal Forming; idle
to analyze CNCcapacity
Machine Tool;
and Control Systems;
to design Kinematics;
strategies towardsSystem
theModelling
maximization of organization’s
1. Introduction production cost are used in automotive, aircraft,
value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity
electronics appliances and in similar manufacturing
optimization
1. Introduction might hide operational inefficiency. production costtraditional
are used method
in automotive,
Sheet metal components are fabricated
© 2017 The Authors. Published by Elsevier B.V. at very low industries. In of sheetaircraft,
metal
cost using several sheet metal forming processes. The electronics
forming, suchappliances
as stampingand ain similar
punch and manufacturing
a die set with
Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference
Sheet
appliedmetal
forcecomponents
stresses thearemetal fabricated
beyond at very low
its yield industries.
desired shape In istraditional
used whichmethod of sheet
can produce a partmetal
in a
2017.
cost using
strength severalthe
causing sheet metaltoforming
material processes.
plastically deformThe but forming, such asbystamping
single stroke a punch
stretching sheetandmetal
a die between
set with
applied forceModels;
not to fail.
Keywords: Cost Bystresses
ABC;the
doing so metal
the
TDABC; beyond
sheet
Capacity metal itscanyield
Management; Idle Capacity;desired
be punch-dieshape
Operational set. isThis
usedmethod
Efficiency which to canfabricate
produce parts
a partusing
in a
strength causing the material to plastically
stretched into variety of complex sheets. [1]. A deform but single stroke by stretching sheet metal
sheet metal is suitable for mass production. However, between
not to fail.fraction
substantial By doing so theproduced
of metal sheet metal can be
as strips is punch-die
using the sameset. This
methodmethod
is not to fabricate parts
economically using
viable for
stretched
1. into
Introduction variety of complex
formed into parts which help in reducing the sheets. [1]. A sheet metal is suitable for mass production.
small scale production or in the industry where However,
substantial fraction of metal produced as strips is using the same method is not economically viable for
formed
The into
cost of parts
idle which
capacity is ahelp in
fundamental reducing
2351-9789© 2018 The Authors. Published by Elsevier B.V. the
information for small
companies scale production
and their management or inof the industry
extreme where
importance
in modern under
Peer-review production systems.
responsibility of theIn general,
scientific it is defined
committee as unused capacity or production potential and can be measured
of NAMRI/SME.
in several ways:
2351-9789© 2018 Thetons of production,
Authors. available
Published by Elsevier B.V.hours of manufacturing, etc. The management of the idle capacity
Peer-review underTel.:
* Paulo Afonso. responsibility
+351 253 of the761;
510 scientific committee
fax: +351 253 604of741
NAMRI/SME.
E-mail address: psafonso@dps.uminho.pt

2351-9789 © 2017 The Authors. Published by Elsevier B.V.


Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017.
2351-9789 © 2018 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 46th SME North American Manufacturing Research Conference.
10.1016/j.promfg.2018.07.012
96 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
2 Author name / Procedia Manufacturing 00 (2018) 000–000

customized products are manufactured such as process has become easier [5]. Simulation of this
medical devices and implants. The cost and time physical manufacturing process using virtual machine
needed in designing and fabricating dies for all new tool prior to physical testing helps in cutting down the
parts in such industry makes conventional stamping time and cost related to testing on a physical machine
method more expensive [2]. Hence this has created tool prototype [6,7]. The fabrication and testing of
demand for innovation in cost efficient and flexible physical prototype after each iteration becomes a
sheet metal forming processes. major hurdle in the development of advanced
One alternative method for metal forming which can machine tool [8]. Hence, the simulation technology
solve these issues is incremental sheet forming (ISF) can drastically reduce the cost and time required in
[3]. In this method, a Computerized Numerical the development of a machine tool [9]. Thus, this
Control (CNC) machine tool is used to create paper deals with the development of ISF CNC
deformation at the location where tool and machine tool model, simulation and control of the
undeformed sheet metal meets. The tool creates CNC machine control system. Traditionally, lumped-
permanent deformation to get a specific shape using parameter models have been used for modeling and
the tool movements. Fig. 1 shows the working of simulation of CNC machine tool [10,11]. Modular
incremental sheet metal forming process. A holder is approach has also been applied to reduce the
used to clamp the undeformed sheet and tool moves
differences in set values and simulated results [12-
in X-Y plane with velocity v and deformation takes
14].
place in vertical direction (Z).
CNC machine tool is an advanced mechatronics
system. Hence modeling a system which considers
both the mechanical system and control system is
very essential. This can be achieved by either using a
separate software package for mechanical system
simulation and using another environment for
mechatronic simulation [15,16]. In this paper
multibody mechatronics model of a ISF CNC
machine tool has been developed using SIMULINK
environment. Complete model has been developed by
combining developed models of electrical and
Fig. 1 . Working of Incremental Sheet Metal Forming Technique mechanical systems. This plays a very important role
in the working of the CNC machine tool. Fig. 2
The movements of tool (coordinates with respect shows the schematic diagram of the different steps
to time) which are controlled by CNC machine tool is involved in the development of CNC machine system
known as tool path. This process is found to be 5- mechatronic model.
10% more economical than the conventional method
of operation where customized parts are fabricated
[4]. This method is also very useful for rapid
prototyping and in the initial product designing stage,
as it reduces the time and cost required for fabricating
die for each new change in the same part. However,
there are limitations in this alternate method of sheet
metal forming. The major limitation is geometric
inaccuracy of the final fabricated part due to spring
back [2]. Hence development and testing of the CNC
machine tool for incremental sheet metal forming
becomes vital to improve the process accuracy.
Control system for CNC machine tool plays a very
important role in the geometric accuracy of the
fabricated part.
With improved computer technology, the
simulation of advanced physical manufacturing
Fig. 2. Steps involved in Development of Mechatronics Model of CNC
Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106 97
Author name / Procedia Manufacturing 00 (2018) 000–000 3

The main objectives of this paper include:

 Modelling a cost efficient and flexible The kinematics of CAD model was validated using
Incremental Sheet Metal forming motion study module of SOLIDWORKS. Finally, the
process, involving development of a assembled model is simplified by combining all fixed
multibody mechatronic model of ISF components together and reducing the total number of
CNC machine using SIMULINK components to four. This will prevent creating
 Simulation and design of an efficient unwanted joints in the MATLAB SIMSCAPE
control system with good accuracy using multibody model [17]. SIMSCAPE is a MATLAB
cascade control module which provides simulation environment for
 Selection of best control strategy from multibody 3D mechanical systems such as robotic
the available combinations manipulator and other complicated kinematic
 Stability analysis of the chosen control systems. Fig.4 shows the SIMSCAPE model of a
system CNC machine.
The ‘World Frame’ block provides a motionless
2. Development of three- axis CNC machine tool co-ordinate frame which is predefined in any
mechatronics model multibody system. It provides ground to all frames
present in the entire mechanical system because any
2.1. Computer Aided Design (CAD) model system can have multiple co-ordinate frames.
‘Mechanism configuration’ block is used to define
The tool path required for incremental sheet metal mechanical parameters such as gravity to the whole
forming was generated using CATIA surface
machining option using Z level operation. The sheet
metal design was used to generate the shapes in Fig.3
(a) & (b) and they were used for the tool path
generation.

The three-dimensional CNC machine tool model


was developed and designed on SOLIDWORKS
CAD software package. Different components such
as frame, lead screw and other parts were separately
designed and were assembled to form a complete
CNC machine tool with bed size of 88 cm x 35 cm
with Z axis limit of 56 cm. The dimensions are
related to a HMT CNC vertical milling machine. The
range of each axis for maximum translational travel
limits, were fixed by applying advance mates. This
allows to analyze the design virtually using animation Fig. 4. Simscape Model of CNC Machine
module on SOLIDWORKS.

multibody system. ‘Rigid Transform’ block is used to


fix the 3D transformation between two frames. It is
used to transform coordinates between world frame
and CNC machine coordinate frame. This
transformation block is linked to the first CAD part of
the CNC model that is the main stationary frame of
the CNC machine. The X axis part block is linked
with stationary frame part block through prismatic
joint. This block is used to define the translational
motion between the two parts. They are actuated from
Fig. 3(a) &(b) CAD Models used for Tool Path Planning
98 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
4 Author name / Procedia Manufacturing 00 (2018) 000–000

external signals and also used to sense displacement, designed control system.
torque and forces acting on the joint. Similarly, other A SIMSCAPE model of this mechanism (Fig.6)
two axes Y and Z are linked through other two based on same principle was developed using various
prismatic joints. SIMDRIVELINE blocks. As shown in the Fig.6, the
lead screw is driven by specifying the input velocity
2.2. Lead screw mechanism model (rad/s) which is provided using ‘ideal angular velocity
source’ block which is used here as a source for
Lead screw mechanism is used to translate rotary motion.
rotational motion generated from motors to
translational motion. This mechanism is more precise
than belt and pulley system due to less slipping and
friction [18]. Hence it is used in machine tools such
as incremental sheet metal forming where high
precision is required.
A typical lead screw mechanism is shown in Fig. 5.
In this mechanical system, a screw is attached to the
shaft of the motor through gear train and shaft coupler
with appropriate gear ratio. Nut is fixed in the nut
housing which carries out translational motion when
the screw is allowed to rotate. The kinematics of the Fig. 6. Simscape Model of Lead Screw Mechanism
lead screw is defined by the following equations [19].
(1)
(2)
(3)
(4)
Input (rad)

Where λ is lead angle (or helix angle), d is the pitch


diameter, l is the screw lead, p is the screw pitch, n is
the number of starts, θ is the lead screw rotation, x is
the nut translation and r is the pitch circle radius.

Time (s)
Output (cm)

Fig. 5. Lead Screw Mechanism

Using machine's CAD model and integrating with


simmechnics toolbox allows to accurately consider
the mass distribution of each mechanical component.
The feed drive's mathematical model is
developed using existing Simulink blocks. This Time (s)
provides freedom to increase or decrease the model
complexity based on the application requirements. Fig. 7. (a) Input & (b)Output response of the SIMSCAPE
For this study, stiffness and friction between each Model of Lead Screw Mechanism
moving component were taken into account to reflect
the real dynamic effects of the machine on the
Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106 99
Author name / Procedia Manufacturing 00 (2018) 000–000 5

This motion has to be relative to some fixed point in


space which is defined using ‘mechanical rotational
reference’ block. Here a sine wave function block
provides a constant angular velocity value to the
angular velocity source through ‘Simulink convertor’
block where the unit rad/s is specified to the input
value. Inertia and mass blocks are used to define the
inertia and mass of the lead screw respectively.
Spring block defines the mechanical load against
which lead screw is acting and it is fixed in space
with ‘mechanical transitional reference’ block. The
input source signal and output of the lead screw are Fig. 8. PM Stepper Motor
shown in Fig.7(a)&(b). Screw lead (displacement per
revolution) is set at 0.015 m with friction. All other
parameters for lead screw such as lead angle, friction As shown in Fig. 8, PM Stepper motor contains two
stator coils A and B which are powered alternatively
coefficient, etc. used for this simulation are shown in
using voltage pulses across A+A- and B+B- to
Table. 1. In the complete CNC machine model,
produce magnetic field to attract opposite poles of the
stepper motor provides output as rad/s through gear
magnetic rotator. This alternative magnetic field
box. makes the motor to rotate in one direction. The
Table. 1.Lead Screw Parameters
mathematical model of this motor is based on
following mechanical and electrical equations.
Parameters Values Electrical equations:
Screw lead 0.015 m (5)
Screw helix type Right-handed (6)
Lead angle 15 deg Mechanical Equations:
Acme thread half angle 14.5 deg
Friction coefficient 0.08
Power threshold 0.001 W (7)
Viscous friction coefficient 250 N m/(rad/s)
(8)
Where, Ia and Ib are the currents in phases A and B
2.3. Permanent Magnet Stepper Motor Model respectively (amps), Va and Vb are the voltages on
phases A and B respectively (volts), ω is the rotor
A permanent magnet (PM) stepper motor provides speed (rad/sec), is rotor position (rad), R is the
discrete mechanical rotational movement which is resistance of the phase winding (Ω), L is the self-
different from conventional DC motor which gives inductance of the phase winding (H), Km is the motor
continuous shaft rotation when it is powered. In a PM torque constant (Nm/A), B is the viscous friction
stepper motor, this discrete step increment is (Nms2/rad), J is the rotor inertia (Kg.m2) and TL is
achieved by providing the electrical command pulses the load torque (Nm).
in proper sequence. The speed and the direction of As shown in Fig.9 Simscape model of PM stepper
rotation are directly proportional to the sequence and motor with power and driver system was developed
frequency of these pulses respectively. It has the using several electronics blocks from Simscape
ability to provide the following: library.PM stepper motor block has four inputs A+,
 Full torque at stand still A-, B+ and B-. These inputs define ends of two stator
 Position precision with non-cumulative error coils of the real motor. They are connected to the
from one step to next output side of the ‘Stepper motor driver’ block. This
block generates the voltage pulse sequences to
 Easy and accurate control using electronic
control the motor. A constant 24 volts was supplied
driver circuit without mechanical drives.
through ENA port which provides the triggering
Due to these advantages, stepper motor was selected
input step voltage and to REF port which is input
in developing the CNC machine model.
100 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
6 Author name / Procedia Manufacturing 00 (2018) 000–000

floating reference voltage or to REV port which is Table. PM Stepper Motor parameters
the input voltage that controls motor direction and
REF port. If the voltage at REV port is greater than
the reverse threshold voltage (2.5 V), then the motor Parameters values
will rotate in reverse direction. Rotational motion is Phase winding resistance 0.55 Ohm
obtained on the output side of the Stepper motor
block. This motion is relative to a fixed point in Phase winding inductance 0.0015 H
space which is defined using ‘mechanical rotational
Motor torque constant 0.19 N*m/A
reference’ block. To run the motor in both directions
for this simulation, a direction controller was Full step size 1.8 deg
developed using two-way single phase switch which
Rotor inertia 4.5e-5 kg*m^2
triggers the motor to change the direction using
external signal. The step signal is used which makes Rotor damping 8e-4 N*m/(rad/s)
the motor to run in forward for first 30 seconds and
reverse for the remaining time. Fig. 10 shows the
rotor velocity and position of stepper motor during
this simulation. All PM stepper motor parameters
used for this simulation are shown in Table. 2.

Fig. 10. Velocity and Position of PM Stepper motor


Simscape Model

Fig. 9. PM Stepper Motor Simscape Model

Fig. 10. Velocity and Position of PM Stepper motor


Simscape Model
Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106 101
Author name / Procedia Manufacturing 00 (2018) 000–000 7

3. Controlling of CNC machine tool 3.2. Cascade control

3.1. Proportional–Integral–Derivative Controller Cascade control provides better control


performance over a conventional single loop control.
A typical closed loop control uses feedback data The performance is enhanced by lowering deviation
available from the system to control it. A schematic and integral error for disturbance response [23].
diagram of such a closed loop control system is Unlike single loop control, cascade control has outer
shown in Fig. 11. The PID controller is by far the main loop with other inner loops and uses additional
most dominating form of feedback control system in information to enhance the performance. Hence
use today [20]. PID controller is used which cascade control can be implemented in the model
constantly determines the error between the desired where more than one process variable can be
and the measured values which is available from the determined. A schematic diagram of a typical
sensors. cascade control system is shown below in Fig. 12.
In a CNC machine, the desired value is the Where A(s) and B(s) are two transfer functions based
position of the link and it is sensed using a position on subsystems of a process and X(s) and Y(s) are the
encoder. Finally, controller attempts to correct the controller blocks. Usual approach in cascade control
position based on error using ‘proportional’, is designing the inner loop first which is followed by
‘integral’ and ‘derivative’ coefficients. This error developing the outer loop. The resulting function due
minimization is performed over time by alteration of to inner loop stated by F(s) in Fig. 12 is taken into
the position. Mathematically this controller is based consideration while developing the outer loop.
on the following equation [21].

(9)
X(t) is position at time t, K(p) is proportional gain,
K(i) is integral gain, K(d) is derivative gain and E is
error. PID, a structurally simple and generally
applicable control technique, stems it success largely
Fig. 12. Cascade Control Diagram
from the fact that it just works very well with a
simple and easy to understand structure [22].
To implement PID controller in CNC machine Cascade control using two controllers provides
model, voltage supplied to the PM stepper motor was better dynamic performance. Here controller present
controlled to obtain the desired motion of all the axes. in the outer loop controls the primary variable of the
The feedback position data was provided from the process which is comparatively slow varying variable
sensor port of ‘prismatic joint’ block available in the such as position. The inner loop controller takes the
CNC machine Simscape model. This feedback data outer loop value as a set point. This loop controls
was compared with the desired axes position. Based more rapid varying variable such as velocity.
on the error, the controller can be tuned. Curve generator is a subsystem which provides
the position data for X-Y and Z-axes. Separate curve
equation for X-Y and Z-axis or co-ordinates data
extracted from G-code can be used to generate
complex trajectories. This data is converted into
desired voltage input through cascaded PID
controllers. Here position is controlled in the outer
loop and velocity in the inner loop. The feedback
data for both the loop is obtained from ‘prismatic
joint’ block available in the CNC SIMSCAPE model.
‘Controlled voltage source’ block provides the
variable voltage input to the PM stepper motor
through motor drive which generates rotary motion.
Fig. 11. Closed Loop Control Based on PID Controller
This rotary motion is converted to translational
102 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
8 Author name / Procedia Manufacturing 00 (2018) 000–000

motion through ‘Lead screw’ subsystem. Finally, the of the cascade controller that deals with motion in x-
translational motion of all the three axes produces axis. From Fig.14 (a) it can be observed that the
desired motion of the end effector. ‘3D scope’ blocks controller in the particular tuning condition closely
are used to view the end effector generated trajectory tracks the reference trajectory. This shows that the
and the desired trajectory in 3D space. Different controller is effective. The rise time and settling time
control strategies for inner and outer loops such as are minimized.
PI-P, PID-PI and PID-PID were tried to determine Relative stability can be determined in terms of gain
the best robust control strategy. In P-PI combination margin and phase margin using the bode plot shown
the Proportional controller is applied to the inner loop in Fig. 14 b & c. Gain margin is obtained from the
and Proportional-Integral is applied to the outer loop. magnitude vs. frequency plot. The phase margin can
Similarly, in the PID-PI combination, Proportional- be determined by using the frequency at which the
Integral-Derivative controller is applied to outer loop zero dB axis cuts the magnitude vs. frequency curve
and Proportional-Integral is applied to the inner loop. (gain cross over frequency). The angle corresponding
In the PID-PID combination, Proportional-Integral- to this frequency from the phase vs. frequency plot is
Derivative controller is applied to both loops. used to calculate gain margin.
When PI-P combinations of controllers were used As the gain margin increases, the stability of the
for the entire three axes, the desired output was not system increases. Similarly, greater the phase margin
reached even after tuning. The same thing holds true greater will be the stability If both the gain margin
for other combinations like PID-PID, PID-P and PD- and phase margin are positive, the system is stable. If
P being applied to the entire three axes. Hence an the phase margin is greater than the gain margin also,
attempt was tried with different combination of the system is stable. A system is marginally stable if
controllers for each axis. The final set-up has the both phase margin and gain margin are zero or phase
following strategies for the three different axes as per margin is equal to gain margin.
Table.3. The control strategies for cascade control Stability analysis of all the controllers for
have been chosen by analyzing the robustness and tracking trajectory in each axis was used for tuning
performance of the controllers individually. each individual controller component that constitutes
the cascade control scheme. Table 4 show the tuning
Table 3. Proposed Controller Strategies parameters pertaining to all cascade controllers
Axis Cascade control strategies implemented. The corresponding values of
Inner Loop Outer Loop parameters that determine the robustness of the
x-axis P PD
controller are also included.
It can be observed from the table that the gain
y-axis P PI
margin and phase margin for each controller in the
z-axis P PI control scheme is positive and hence every controller
is stable. Further, the table also shows that the rise
time and settling time are low for all the loops.
4. Results and Discussion This particular configuration of controller and
tuning is very efficient and it perfectly enables
The various combinations described in Table.3. tracking of both the trajectories. Fig. 15 (a-b) shows
have been implemented in the inner and outer loop of simulation results of closed Loop responses of
the cascade control of the three axes and Fig.13 reference input and output when the trajectory is
shows the SIMULINK block diagram which was trapezoidal whereas Fig. 15(c-d) shows input and
finally implemented output responses when the trajectory is spherical.
For the x-axis, the outer-loop controller is most Cascade control system also has to ensure good
robust when it is a PD controller tuned to P = disturbance rejection. The correct operation of both
477.4210 and D =2.1029. At this particular tuning cascade control system and feed-forward controls of
condition, the reference tracking is most efficient and a servo-drives improves the accuracy of a machine
the controller is stable. Fig. 14 (a) shows the tool operation [24].
reference tracking while Fig14(b&c) shows bode plot
which can be used for the analysis of the outer loop
Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106 103
Author name / Procedia Manufacturing 00 (2018) 000–000 9

Fig. 13. Complete Closed Loop Model of CNC under Cascade Control

Fig.14. (a) Sample Reference Tracking; (b) & (c) Sample Bode Plots
(Magnitude and Phase Plots) for CNC Machine Model
104 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
10 Author name / Procedia Manufacturing 00 (2018) 000–000

Table 4. Tuning Parameters and Performance Parameters that


determine Stability of the Control

Loop Tuning Corresponding performance parameters


Parameters
Parameter Tuned Block

x Inner P 0.79 Rise time Nano sec Nano sec


895 Settling time Nano sec Nano sec
I NA Overshoot NaN% NaN%
Peak Inf Inf
D NA Gain margin 7.48dD@873 5.53dD@873
rad/s rad/s
Phase margin 57.2deg@441 40.9deg@557
rad/s rad/s
Closed loop Stable Stable
stability
Outer P 477. Rise time Nano sec Nano sec
421 Settling time Nano sec Nano sec
0
I NA Overshoot NaN% NaN%
Peak Inf Inf
D 2.10 Gain margin 9.13dD@477 7.93dD@873
29 rad/s rad/s
Phase margin 6.15deg@189 6.12deg@413
rad/s rad/s
Closed loop Stable Stable
stability
y Inner P 0.79 Rise time 0.346 seconds 0.38 seconds
895 Settling time 2.06 seconds 2.17 seconds
I NA Overshoot 20.7% 22.3%
Peak 1.21 1.22
D NA Gain margin 37.1dB@255 38.2dB@255
rad/s rad/s
Phase margin 64.9deg@3.7 62.6deg@3.3
3 4
rad/s rad/s Fig. 15. (a-b) Simulation Results for Closed Loop Responses of
Closed loop Stable Stable (a)Reference Input and (b) Output when the Trajectory is
stability
Outer P 0.47 Rise time 4.15 seconds Nano sec Trapezoidal
585 Settling time 34.3 seconds Nano sec
I 0.03 Overshoot 13% NaN%
217 Peak 1.13 Inf
4
D 0 Gain margin 65.8Db@412 19.4Db@474
rad/s rad/s
Phase margin 72deg@0.355 69.9deg@0.8
rad/s 34
rad/s
Closed loop Stable Stable
stability
z Inner P 1.14 Rise time Nano sec Nano sec
057 Settling time Nano sec Nano sec
I NA Overshoot NaN% NaN%
Peak Inf Inf
D NA Gain margin 7.48dB@873 5.53dB@87
rad/s 3rad/s
Phase margin 57.2deg@441 40.9deg@557
rad/s rad/s
Closed loop Stable Stable .
stability
Outer P 1.14 Rise time Nano sec 0.0757
057 seconds
Settling time Nano sec 0.801seconds
I 6.54 Overshoot NaN% 3.41%
518 Peak Inf 1.03
D NA Gain margin 15.5dB@369r 27.6dB@369r
ad/s ad/s
Phase margin 62.6deg@93. 80.7deg@24.
9rad/s 5rad/s
Closed loop Stable Stable
stability

Fig. 15. (c-d) Reference Input and (d) Output when the
Trajectory is Spherical Cup
Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106 105
Author name / Procedia Manufacturing 00 (2018) 000–000 11

There are geometric inaccuracies associated with disturbance will help to determine the most robust
incremental sheet metal processes due to spring back control strategy. Table 5 illustrates the values of ISE
as a result of elastic recovery of the sheet metal. The and IAE errors for two different predefined
surface finish may be affected by chatter and trajectories as well as simulation time for the
vibrations caused during incremental sheet metal trajectories being tracked in the presence and absence
forming. Further there may be noise in the feedback of disturbances. It is observed from the Table 5 that
loop contributed by the inaccuracies from sensors. even in the presence of disturbances in both the loops
Stability analysis essentially deals with disturbances all ISE, IAE and simulation times are almost same.
that arise from sensors etc. directly affecting the It can be observed that there isn’t a significant
position readings. increase in ISE and IAE even with the presence of
The given step disturbance has a magnitude of one. disturbance.
In the proposed method, the occurrence of an abrupt Table. 5. Performance Analysis for Various Trajectories with and
(namely, step-like) load disturbance has been Without Disturbances
DISTRURB SIMULA
assumed. Indeed, this is the most relevant case for the SHAPES
DISTURBA
NCE –
ANCE –
OUTER IAE ISE
TION
TIME
control system, as the disturbance excites INNER
LOOP
LOOP (s)

significantly the dynamics of the control system itself


Trapezoidal Absent Absent 3272 4.72e04 586
[25]. After the abrupt disturbance, the trajectory is cup
traced accurately again. Hence, our controllers are Spherical cup Absent Absent 2298 2273 2994
Trapezoidal Absent Present 630.8 822.5 588
robust and give output even when disturbance is cup
introduced. Fig. 16 shows the trajectory tracked when Spherical cup Absent Present 2301 2278 2994
Trapezoidal Present Present 3278 4.75e04 583
a spherical cup input and step disturbance are given. cup
Spherical cup Present Present 2353 2329 2994

The PID controller is widely used in industrial


control and achieves good control effect due to its
simple structure and easy control [26]. At present
controlling sheet metal process using PID controllers
in cascade is well suited for cost effective
manufacturing.
To further enhance the results, an adaptive
optimization and PID control system presented for
optimizing a multi-objective function and control of
servo drive systems can be used [26]. Fuzzy PID
controller may be used as proposed by Xie Dong
Fig.16. Output Response of Cascade Control for a Step Change
in Disturbance in the Inner and Outer Loops [27]. Hence, the future scope of this paper includes
using optimization of the motor parameters as well as
replacing the conventional PID controllers with
Further the performance of the cascade control was Fuzzy PID.
also analyzed by determining Integral Square Error
(ISE) and Integral Absolute Error (IAE) for the 5. Conclusion
defined end effector positions with and without
disturbance. The Integral Squared Error is defined as: The detailed procedure for modeling and simulation
(10) of ISF process using CNC machine tool has been
presented. The cascade control has been designed and
The Integral Absolute Error is defined as:
tested through simulation to control the velocity and
(11) position of the machine tool to obtain the desired
trajectories. Trapezoidal and spherical cup
Where ε is error in end effector position. These errors trajectories generated using CATIA software are used
were measured over different predefined trajectories for testing in the SIMULINK model developed. The
namely trapezoidal cup and spherical cup. Analyzing methodology can be extended to hardware-in-loop
these errors for the two trajectories and also including experiment along with velocity and position sensors
106 Ketul Patel et al. / Procedia Manufacturing 26 (2018) 95–106
12 Author name / Procedia Manufacturing 00 (2018) 000–000

to validate the results in future. The disturbance in international intelligent motion conference, (1999) pp.259-
the model is simulated to represent the spring back 416.
[13] Erkorkmaz, K., Altintas, Y., and Yeung, C. ‘Virtual computer
during sheet metal forming which is the main source numerical control system’ CIRP Annals-Manufacturing
of inaccuracy. Hence developing, performing Technology, Vol. 55 No.1,(2006) pp.399-402.
complex analysis and optimizing the system using [14] Zaeh, M.F., Oertli, T., and Milberg, J. ‘Finite element
different strategies can significantly reduce the cost modelling of ball screw feed drive systems’ CIRP Annals-
and time needed in developing a tool machine with Manufacturing Technology, Vol. 53 No.1,(2004) pp.289-292.
[15] Sztendel, S., Pislaru, C., Longstaff, A. P., Fletcher, S., and
every iteration in the design. Virtual simulation also Myers, A. ‘Five-axis machine tool condition monitoring using
enables the effective control of position accuracy and dSPACE real-time system’ Journal of Physics: Conference
control dynamics of a machine tool. Series, Vol. 364, No. 1 (2012).
[16] Neugebauer, R., Scheffler, C., and Wabner, M.
Acknowledgements ‘Implementation of control elements in FEM calculations of
machine tools’ CIRP Journal of Manufacturing Science and
Technology, Vol. 4 No.1,(2011) pp.71-79.
The authors are grateful to the authorities of BITS [17] Papananias, M., Sztendel, S., and Pislaru, C. ‘Development of
Pilani, Dubai Campus for their support to carry out a Novel MultiBody Mechatronic Model for Five-Axis CNC
this research work through seed grant. Machine Tool’ EUSPEN,(2015) pp.379-389.
[18] Varanasi, K and Nayfeh, S. ‘The dynamics of lead-screw
drives: low-order modeling and experiments’ Transactions-
References American Society of Mechanical Engineers Journal of
Dynamic Systems Measurement and Control, Vol. 126 No.
[1] Hu, J., Marciniak, Z., & Duncan, J. (Eds.). Mechanics of sheet 2,(2004) pp.388-396.
metal forming, Butterworth-Heinemann. (2002). [19]Vahid-Araghi, O. and Golnaraghi, F. ‘Friction-induced
[2] Hao, W., and Duncan, S. ‘Optimization of tool trajectory for vibration in lead screw drives’, Springer Science & Business
Incremental Sheet Forming using closed loop control’, Media.(2010).
Conference on Automation Science and Engineering [20] Guzman J L, Astrom K J, Dormido S, et al. ‘Interactive
(CASE),(2011) pp.779-784. learning modules for PID control’ [Lecture Notes][J]. Control
[3] Jeswiet, J., Micari, F., Hirt, G., Bramley, A., Duflou, J., and Systems, IEEE, (2008) pp.28(5): 118-134.
Allwood, J. ‘Asymmetric single point incremental forming of [21] Patel Ankur, Vijay Savani, ‘Performance Analysis of PID
sheet metal’, CIRP Annals-Manufacturing Technology, Vol 54 Controller and Its significance for Closed Loop System’,
No. 2,(2005) pp.88-114. International Journal of Engineering Research & Technology
[4] Owodunni, O., Hinduja, S., and Mekid, S. ‘Towards rapid sheet (IJERT), Vol. 3 Issue 3, (2014)pp.1843-1847.
metal forming’ Proceedings of 1st I* PROMS Virtual [22] Kiam Heong Ang, Gregory Chong, and Yun Li, Member, PID
International Conference on Intelligent Production Machines Control System Analysis, Design, and Technology, IEEE
and Systems, IPROMS,(2005) pp. 573-578. Transactions on Control Systems Technology , Vol. 13, No.
[5] Pandilov, Z., Milecki, A., Nowak, A., Górski, F., Grajewski, 4,( 2005) pp. 559 – 576.
D., Ciglar, D., and Mulc, T. ‘Virtual Modelling and simulation [23]Maffezzoni, C., Schiavoni, N., and Ferretti, G ‘Robust design
of a CNC machine feed drive system’ Transactions of of cascade control’ IEEE Control Systems Magazine, Vol. 10
FAMENA, Vol. 39 No. 4,(2016) pp.37-54. No.1,. (1990) pp.21-25.
[6] Owen, J. V. ‘Virtual manufacturing’, Manufacturing [24] S. Mandra, ‘Comparison of Automatically Tuned Cascade
Engineering, Vol. 119 No. 4,(1997) pp.84 Control Systems of Servo-Drives for Numerically Controlled
[7] Radharamanan, R. ‘Virtual manufacturing: an emerging Machine Tools’, ELEKTRONIKA IR
technology’ Proceedings of the 2002 American Society for ELEKTROTECHNIKA, ISSN 1392-1215, Vol. 20, No. 3
Engineering Education Annual Conference & (2014).
Exposition(2002)pp.1. [25]Massimiliano Veronesi ,Antonio Visioli, ‘Performance
[8] Weck, M., Queins, M., and Witt, S. ‘Effektive Entwicklung Assessment and Retuning of PID Controllers for Load
von Werkzeugmaschinen’ VDI-Z, Vol. 145 No.10, (2003) Disturbance Rejection’, IFAC Conference on Advances in PID
pp.32-36. Control, PID'12 (2012).
[9] Altintas, Y., Brecher, C., Weck, M., and Witt, S. ‘Virtual [26] Susai Mary ,J Sabura Banu U, Dinakaran D, Nakandhrakumar
machine tool’ CIRP Annals-manufacturing technology, Vol. R S, ‘Adaptive Optimization using Grey Relational Analysis
54 No. 2, (2005) pp.115-138. and PID Control of CNC Drilling Process’, International
[10] Leonhard, W. Control of Electrical Drives, Springer, Conference on Robotics, Automation, Control and Embedded
Germany.3rd ed(2001). Systems .,DOI:10.1109/RACE.2015.7097277(2015).
[11] Erkorkmaz, K., and Altintas, Y. ‘High speed CNC system [27] Xie Dong, Zhu Jian-qu, Wang Feng, ‘Fuzzy PID Control To
design. Part II: modeling and identification of feed drives’ Feed Servo System of CNC Machine Tool’, International
International Journal of Machine Tools and Manufacture, Vol. Workshop on Information and Electronics Engineering
41 No.10,(2001) pp.1487-1509. IWIEE , Procedia Engineering 29 (2012) pp.2853 – 2858
[12] Pislaru, C., Ford, D., and Freeman, M. ‘Dynamic simulation
of CNC machine tool axis drives’ Proceedings of the

You might also like