Development , Kinematic Modeling and Analysis of 3 ( DoF ) Pneumatic Gantry Robot

DOI: http://dx.doi.org/10.24018/ejers.2020.5.6.1917 Vol 5 | Issue 6 | June 2020 646 Abstract—Low friction pneumatic cylinders are now being considered in applications for which only electric motors or hydraulics were previously considered suitable. One potential application of low friction pneumatics is robotic for metallurgical operations where the high power to weight ratio and low cost could be exploited. As part of an ongoing project to develop a pneumatic robot, this paper presents the kinematic analysis of pneumatic cylinder characteristics that simplifies controller design. Using mathematical modeling and simulation, non-linearity of modern pneumatic systems has been investigated. The derived models give an excellent representation of the system, despite the inclusion of a simplified friction model.


I. INTRODUCTION
Robot kinematics is mainly falling in to two types: forward and inverse kinematics. Forward kinematics is also known as direct kinematics. In forward kinematics, the length of each link and the angle of each joint are given, and we have to calculate the position of any point in the work volume of the robot [1]. In inverse kinematics, the length of each link and position of the point in work volume is given and we have to calculate the angle of each joint. To accomplish a task in an application, we need to control the position and orientation in various coordinate systems such as world, work piece to tool. In order for a robot to go to certain place at certain orientation conveniently, it is necessary to know the relationship between the joint coordinate system and some other systems, such as base or tool systems [2]. Robots are used in many industries like automobile, semiconductor, electronics, aerospace food and beverage, pharmaceutical, consumer goods, plastics, construction, and medical devices. However, the use of robots for metallurgical operations such as hot-metal ladle handling, heat treatment etc. is rather limited due to many environmental and safety reasons [3] [4].
In the present work, the development of foreword kinematic modeling and simulation for the pneumatic robot (scaled-down model) is described in detail. Cartesian type manipulator, b) each degree of operation by rod-less and double acting pneumatic actuators, c) Air Operated vacuum gripper, d) Capacitive proximity sensor for detecting the presence of the bucket containing components for heat treatment and e) Two types of operational convenience, pushbutton for continuous operation and on/of, switch for one cycle operation. The function of the above Pick & Place robot is: a) to detect the presence of the bucket in front of the capacitive proximity sensor and b) to pick the basket of components to be heat treated/ ladle for pouring, and place it in the furnace / to position over moulds in a gentle manner. Fig. 1 shows the constructional details of the 1:8 model of the proposed robotic system for metallurgical operations. This scheme is proposed to be extended for threedimensional robot later.  (DoF) Pneumatic Gantry Robot containing components to be heat treated, station 2 corresponding to a heating oven for quenching or tempering, station 3 for the vat of quenching medium or a sand bed for slow cooling and station 4 for unloading of the buckets on an output conveyor. The manipulator transformation matrix and D-H representation [5].

II. PNEUMATIC ROBOT ARM FOR PICK AND PLACE
The transformation matrix at every joint is given by Danavit -Hardenberg representation After using the above transformations, Therefore, the final transformed matrix is given by Hence, • XE = a2, YE = d3, ZE = d1 The end-effecter coordinators are = (a2, d3, d1) and in the present case are: a2 = 200 mm, d3=500 mm, d1 = 30 mm at the end of the operation of the pneumatic cylinder.

B. Linear Velocity of End-Effecter (V)
By using Jacobean relationship we can found Endeffecter velocity as fellow: We note that this is Jacobean matrix it has no singular points and the robot can move into any point of its workspace. Then:

C. Forward Kinematic Gantry Robot System-Study & Result
The robot performance is related to the specification of its tasks. A task can be interpreted as a temporal relation between kinematics and inertial parameters. After the definition of the manipulator movement, the kinematics parameters are obtained for the specified path. Then these data are solved by MATLAB, which results are used to determine the end-effectors extremity position and velocity.  The control system is designed based on PLC instead of human power. The development of this system can be eliminated the usage of human power to pick up things, increase productivity, efficiency, pinch the production time, Rapid, flexible and safe by experiments [7][8].
The control system is configured by using a programmable logic controller as a controller as well as magnetic proximity sensors for sensing purpose and solenoid direction control valves for actuation.
The control scheme is given in Fig.6 [9][10]. The simulation of the robot scheme has been carried out using automation studio software. [3] In these simulations, the rod less cylinder is operated with a pressure of 5 bar. Figs.7-10 show variation of a) linear speed of rodless, b) vertical cylinder for picking up the bucket, c) gripper cylinder and d) acceleration of the rodless cylinder.

VII. CONCLUSIONS
A three-dimensional pneumatic robot is produced at first which is amplified later into an undeniable robot for utilization in a metallurgical application. The general conduct of the reenactment nearly looks like that of the real framework. The modest PLC controller has ended up being sufficient for utilization in such application.
Simulation is done by means of Automation Studio, version 5.6 software. Automation Studio is a system design, simulation, and documentation software solution for the design and support of automation, hydraulics, pneumatics and electrical systems. Parameters of position, velocity, acceleration, pressure variation and flow are obtained from the simulations and compared with those obtained from a video analysis and physical measurements.
The measured values of position and velocity obtained from computer simulation are compared with those obtained through measurements and found to correlate fairly well. The discrepancies in results between simulation and performance are due to friction, which is not truly accounted in the simulation process.