See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/251919938

Structural analysis of a delta-type parallel industrial robot using flexible

dynamic of ANSYS 11.0

Article · November 2009

DOI: 10.1109/IECON.2009.5415370

CITATIONS READS
6 1,877

2 authors, including:

Luis Angel
Pontifical Bolivarian University
74 PUBLICATIONS 706 CITATIONS

SEE PROFILE

All content following this page was uploaded by Luis Angel on 26 February 2014.

The user has requested enhancement of the downloaded file.

Structural Analysis of a Delta-Type Parallel Industrial
Robot Using Flexible Dynamic of ANSYS 11.0
J. D. Rueda
Bachelor degree in Mechanical Engineering and Electronic Engineering
E-mail: [email protected]

L. Ángel
Robotics and Vision Group (RoVi), Electronics Engineering Faculty
Universidad Pontificia Bolivariana, Bucaramanga, Colombia
E-mail: [email protected]

Abstract- This paper presents a novel simulation tool for

structural analysis of a delta robot using SOLID EDGE and
II. SIMULATION MODELS
ANSYS 11 that allows optimization of the robot’s mechanical The simulation models used two simulation software
structure as regards to materials, geometry and manufacture programs: a computer-assisted design program named
costs. To develop this simulation tool, a three-dimensional model
SOLID EDGE and a finite element analysis program
of the delta robot was obtained through SOLID EDGE, and
trajectory analysis was carried out, resulting in selection of
named ANSYS 11. These two simulation programs work
critical trajectory between several possible movements. Finally, together because of their special features, ideal for design
dynamic structural analysis of the robot was run on Flexible and analysis of mobile structures like robots, allowing
Dynamic ANSYS 11, taking into account the dynamic forces of simulation of engineering designs, assessment of their
inertia intervening in the model, as well as stress analysis along features, properties, feasibility, and profitability, and also
the selected critical trajectory. optimization of their development and subsequent
production costs. Simulation models on Solid Edge and
I. INTRODUCTION ANSYS 11.0 are shown in Fig. 2 and Fig. 3.
In general, the design process of an industrial robot is
determined by specific operation features called for by
application needs. Some operation specifications are:
maximum work space, top speed of final effector, positioning
accuracy, structure weight, useful life, materials used, and
production costs. Once operation specifications are defined,
the design process of an industrial robot falls in the following
stages: kinematic optimization, trajectories analysis, dynamic
optimization, and control system design. Kinematic
optimization aims to select mechanical structure and length of
each one of its links or segments. Trajectories analysis defines
possible “critical trajectories”. It is in these trajectories that the
robot’s mechanical structure is subject to maximum stress. Fig. 1. Simulation model on Solid Edge
Dynamic optimization uses dynamic structural analysis in
defining the strains, stresses, and deformations that affect the
structure, so as to optimize materials, geometry, and mass of
each mechanical structure component, and to commercially
select the robot’s actuators. Control system design takes into
account the system’s dynamics, and it allows the effector to
meet specific operation requirements in terms of precision,
speed and movement smoothness [1].
The main objective of this document is the analysis of
trajectories and dynamic optimization of a delta-type parallel
robot, based on dynamic structural analysis [2]. This analysis
enables optimization of the robot as to materials, geometry,
mass, and selection of actuators. The result is more economic
and efficient structures that will meet desired operation
specifications. Fig. 2. Simulation model on ANSYS 11

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2247

 The DELTA robot consists of a moving platform (2) maximum stress analysis. This trajectory will serve as the
connected to a fixed base (1) through three parallel kinematic basis for structural analysis of the delta robot and
chains(3-4). Each chain contains a rotational joint activated by optimization of its mechanical structure.
actuators (5) in the base platform. The motion is transmitted to Trajectories analysis was conducted taking into
the mobile platform through parallelograms formed by links consideration the specifications of the maker of Delta IRB
and spherical joints. 340 FlexPicker robot.
This project utilizes as the basis for the simulation model the • Workspace: (1130 mm) in diameter.
dimensions and operation features of Delta IRB 340 • Top speed: (10 m/s).
FlexPicker robot (Fig. 3). This international-standard industrial Trajectories for the delta robot were developed using a
robot is made by ABB. trajectory planner based on polynomials 616 [4], with a
speed of (2 m/s).

A. Selection of Critical Trajectory

Three possible critical trajectories were taken for this
study. They were loaded on Flexible Dynamic of ANSYS
11.0, in order to find the trajectory that will produce the
maximum stress on the robot’s mechanical structure. For
every movement analyzed, Fig. 5, 6, and 7, the workspace
utilized corresponds to that of a cylinder with a 500 mm of
diameter and 250 mm of height. Effector speed is (2 m/s).

Fig. 3. Delta IRB 340 FlexPicker robot [3].

A. Flexible Dynamic Simulation Model

The tool Flexible Dynamic de ANSYS 11.0 allows
definition of the loads intervening in the dynamic analysis of
the simulation model. In the case of the delta robot, the loads
are the angular displacements of the three motors and the force
of gravity.
The complete simulation model and the ANSYS 11.0
dynamic structure setting are shown in Fig. 4. Angular
movements are shown by means of arrows turning around each
motor and the force of gravity is represented by its direction.
Fig. 5. ”Star” movement 1 on 3D.

Fig. 4. Complete dynamic structural setting.

III. TRAJECTORIES ANALYSIS

Trajectories analysis involves the study of different types of
critical movements that the final effector of the delta robot is
Fig. 6. “Cross up and down” movement 2 on 3D.
required to develop, and selection of critical trajectory through

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2248

 Fig. 10. “Square down” movement 3 strain
(Stress vs. Time).

As observable from results shown above, movement 1

brought about maximum stress of (18.39 Mpa), movement
Fig. 7. “Square down” movement 3 on 3D. 2 brought about maximum stress of (16.73 Mpa) and in
turn, movement 3 brought about maximum stress of
Figures 8, 9 and 10 show the stress involved in each movement (25.271 Mpa). Given the maximum level of movement 3,
of the test. movement 3 is selected as the simulation model’s critical
trajectory.
Figure 11 shows a sequence of the movements produced
by the robot in the development of the critical trajectory
selected. It shows the different positions adopted by the
robot’s mechanical structure along the development of the
trajectory.

Fig. 8. “Star” movement 1 strain (Stress vs. Time).

Fig. 9. “Cross up down” movement 2 strain (Stress vs. Time). Fig. 11. Critical trajectory movement carried out by the Delta robot on
ANSYS 11.0

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2249

 IV. STRUCTURAL ANALYSIS AND OPTIMIZATION OF THE B. Aluminum Vs. Structural Steel
MECHANICAL STRUCTURE Figures 12 and 13 show that geometry of the simulation
This section presents the methodology followed in model is the same for both materials, with a volume of
conducting the structural analysis of the Delta robot using a top 1.6389E6 mm3. However, masses vary due to difference in
speed of (10 m/s). This methodology uses the critical trajectory the densities of the materials.
selected and the simulation model developed on ANSYS 11.0. Stress analysis on Flexible Dynamic Ansys 11.0 shows
Maximum stress and the safety factor are used as the maximum stress of (42.643 Mpa) and (18.59 Mpa) for steel
evaluation criteria of the mechanical structure. The critical- and aluminum arms, respectively, (Fig. 14). It is shown that
piece selection process was initially carried out in the for dynamic cases where the force of inertia is a factor,
simulation model. This simplifies structural analysis and higher mass levels or higher mechanical (static) properties
allows work on the piece that undergoes the most stress. Then, will constitute no guarantee of higher mechanical resistance
the construction material of the mechanical structure is in a dynamic analysis, the following theory: In a dynamic
selected. Finally, the mass of the piece undergoing the most machine, adding weight (mass) to moving parts could have
stress, and made from previously selected material, is the opposite effect, reducing the safety of the machine, its
optimized. permissible speed, and its payload capacity. This arises
from the fact that part of the load that generates stress on
A. Selection of Materials moving parts is the consequence of the inertia laws
In the design methodology proposed, selection of materials foreseen in Newton’s second law, F=m.a. Given that
will be conducted by varying the construction material of the acceleration of moving parts is determined by their
critical piece and by means of stress analysis, so as to find the kinematic design and their operation speed, adding mass to
type of material that most efficiently meets operation moving parts will increase the loads from inertia on those
specifications. For this analysis, aluminum and structural steel same parts or pieces, unless their kinematic acceleration is
are used as study materials. reduced by slowing down operation. Even though the mass
Figures 12 and 13 show the mechanical characteristics of the added might increase resistance of the pieces, such benefit
arm made of aluminum and structural steel. It is observable would be reduced or cancelled by resulting increases from
that the arm made of aluminum has a mass of (4,5397 kg)., the force of inertia “becoming the victims of their own
whereas the arm made of steel has a mass of (12,865 kg). mass”.

Fig.12. Mechanical characteristics of the arm made of aluminum

(a)

Fig.13. Mechanical characteristics of the arm made of structural steel (b)

Fig. 14. Graphs (Stress (Mpa) vs. Time(s)):
(a) Aluminum, (b) Structural steel.

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2250

 The previous analysis, and the fact that the movement speed
considered for the analysis is (10 m/s), lead to the conclusion
that the larger the mass in motion, the larger the par (pair),
involving a larger, costlier motor, which makes aluminum the
material with the best performance for the delta robot, and the
one chosen for the project, as well as the basis for the
following optimization step.

C. Optimization of Mass
Optimization of mass is finding the adequate mass for a
piece, taking as the design evaluation criterion the safety factor
for a specific number of work cycles.
Delta IRB 340 FlexPicker (ABB) robot was previously taken
as the reference, and it can conduct an average 150 picks per
minute, or 788400000 work cycles along 10 years. For the
proposed simulation model, a 700- million work cycle in 10
years and a safety factor (SF) of 1 have been used. Fig. 15. Model 1, arm mass equals 4.5397 Kg., stress produced of
(18.59 Mpa), FS > 15.

1. Simulation models with different masses

Mass optimization is an iterative process. The iterative
process starts out with a determined mass, and the maximum
stress and safety factor are pinpointed. If the safety factor
exceeds desired levels, mass is eliminated, and the safety
factor is pinpointed once more. This process is repeated until
the desired safety factor is reached.
Figure 15 shows the stress undergone by the aluminum arm
when the critical trajectory is followed. In this first model, the
total mass of the arm (4,5397 Kg.) is taken. Given that strain is
too low and the safety factor exceeds 15, further iteration and
reduction of mass were necessary. As an example, Figures 16
and 17 shown two iteration models where mass has been
reduced. The total number of iteration models used was 8.
The relation between mass variation and safety factor, as
well as mass variation and strain obtained in each step of the
iterative process are shown in Figures 18 and 19. there is a Fig. 16. Model 6, arm mass equals (1.5771 Kg)., with a strain of
minimal strain increase when the arm’s mass varies between (53.576 Mpa), FS =7.608.
(4.5 kg) and (1.7235 Kg). Below (1.7235 kg), a minimal mass
reduction had a significant increase in strain. This agrees with
the safety factor results, where models 1 through 5 represent a
safety factor in excess of 15, and only ulterior models show
significant safety factors. It can also be concluded that in order
to optimize mass, it is necessary to reduce mass wherever the
structure is over-dimensioned (zone of minimal strain), and
add mass wherever it is necessary (zone of maximum strain).
For this case, in the first iteration it was found that the arms
were over-dimensioned, given that the safety factor is > 15.
This was the piece selected for the mass optimization iterative
process. Theoretically, a machine that poses no risk for human
life, like the Delta robot, is considered over-dimensioned if the
safety factor exceeds 3. For the simulation model proposed, a
safety factor = 1 has been selected.
As a conclusion, it can be stated that the model that met the
design criteria (700 million cycles with a safety factor below 3)
was model 8. For this model, the SF = 1.07.
Fig. 17. Model 8, arm mass equals (1.3083 Kg)., with a strain of
(120.04 Mpa) FS =1.073.

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2251

 of construction, and service, among others. For the
simulation model proposed in this project, the evaluation
criterion used was the safety factor for a number of 700
million work cycles. Implementation costs were
additionally taken into account.
Tasks proposed for the future include: extending
structural analysis of serial-type and parallel robots of
whatever degree of freedom, since the methodology
developed for this delta robot is usable for whatever robot
type.

Fig. 18. Mass-(Kg) vs Safety Factor relation for all models.

Fig. 19. Mass-(Kg) vs. Strain (Mpa) relation for all models.
Fig. 20. Moments for model without optimization (Mass 4.5397 Kg.).
D. Selection of Motors.
Figures 20 and 21 show the torques for the motors according
to the initial model (without mass optimization), and according
to the final model (with the minimal mass possible),
respectively. These figures show the moments necessary in the
motors to comply with the critical trajectory.
A maximum torque of (337 Nm) can be observed for the
simulation model without optimization process, and a
maximum torque of (165 Nm) for the optimized simulation
model. These results show that the optimization methodology
proposed for the robot obtains the torques necessary for each
motor, guaranteeing compliance with the operation conditions
initially thought up for the mechanical structure. Additionally,
this tool allows commercial selection of the motors, showing
on the implementation costs of the final prototype.

Fig. 21. Moments for optimized model (Mass 1.3083 Kg.).

V. CONCLUSIONS
In this paper, a methodology for structural dynamic analysis
of a parallel Delta robot using ANSYS 11 simulation software VI. REFERENCES
has been developed. The methodology allows optimization of
[1] Mott, R. L. (2006). Machine element design (4th ed.) Prentice Hall.
the robot’s mechanical structure in terms of materials, [2] Clavel, R. DELTA: a fast robot with parallel geometry: 18th
geometry and manufacture costs. International Symposium on Industrial Robot. pp. 91-100. Sydney,
Therefore, this methodology allows selection of the best motor, Australia, 1988.
with the lowest power consumption, the most efficiency in the [3] ABB, Ltd. Swot Analysis. MA : EBSCO Publishing, 1999.
[4] Angel, L. Control visual de robots paralelos. Análisis, desarrollo y
weight-potency ratio, and thus lowers costs. Dynamic analysis aplicación a la plataforma robotenis”, Tesis Doctoral, Escuela Técnica
allows visualization of the distribution of strain of each Superior de Ingenieros Industriales, Madrid, España, 2005.
position along the entire trajectory, which makes this project
novel and meaningful, because distribution of strain used to be
possible for only one position of the whole trajectory.
Testing of the structure can be conducted under evaluation
criteria that use maximum strain, useful life, safety factor, ease

978-1-4244-4649-0/09/$25.00 ©2009 IEEE 2252

View publication stats