4.DISCUSSION OF THE EXPERIMENT
The various degrees of freedom were tested inpidually and then run together as a complete unit using the developed GUI. The tool changer was able to deliver on the required specifications, but there were some limitations that needed to be addressed in order to improve the unit’s capabilities and performance. A successful tool change was performed in less than the required 30 seconds.
The first important limitation was the Festo linear actuators. They had to be programmed to a set position table before integration with the rest of the system. This meant that the various positions of the gripper relative to the machine it was interacting with had to be known ahead of time – or that the machine would be in down-time while the tool-changing positions were taught and programmed in. Once programmed, the Festo drives, along with their corresponding controllers, proved to be a very accurate source of motion. The repeatability and accuracy of the drives was 0.2mm.
The rotation about the horizontal axis also revealed some complications. The torque demands on the motor varied as the angle was changed. This resulted in an accuracy of only 2 degrees being realised for this degree of freedom.
In terms of accuracy, the rotating carousel yielded good results. Low speeds were required to avoid gear kickback, but the carousel was accurate to within 1.5mm along the circumference. It also displayed a repeatability of 1mm. The initial design specifications of the unit were that it could carry six tool holders. In the end, the carousel contained eight slots with room for several more should the need arise.
5.ENABLING TECHNOLOGIES FOR AUTOMATIC INTEGRATION OF THE TOOL CHANGER INTO THE MANUFACTURING SYSTEM
In order for the unit to deliver automatic tool-changing in a reconfigurable manufacturing system environment, the core RMS characteristic of integrability with respect to the tool changer needs to be addressed. One method of applying this characteristic practically is to enable the unit to have an automatic calibration capability. Further research will investigate the tool changer’s ability to exhibit diagnostic features, thereby improving its diagnosability.
5.1 Calibration and integrability
Integrability defines the efficiency with which different components in a reconfigurable system can be added to each other or to the system as a whole. Its definition, according to Mehrabi [8] and ElMaraghy [9], implies integration with the current system as well as an ability to be integrated into future systems with future technologies. Therefore, for a higher level of integrability, one not only has to consider the current configuration of a system: future configurations also have to be taken into account. Designs of reconfigurable machines need to be forward-thinking, and highly reconfigurable modules will be able to adapt to future changes without having to be rebuilt or redesigned.
Integrability affects the time required to reconfigure a machine, which in turn affects the life-cycle cost of the equipment [2]. It is concerned with both the hardware and the software components of a module [1].
The tool changer is a mechatronic system. There are two primary factors that affect the accuracy of mechatronic tools. They are:a) the precision of the mechanical components;b) the ability of the software and control architecture to adjust and correct errors [11].
The tool changer is required to form part of a reconfigurable manufacturing system – a system designed to be changed and adapted to suit different manufacturing demands. It can therefore be seen that the level of integrability of the unit will depend largely on its ability to be efficiently re-calibrated to a different configuration or machine.
A typical robot calibration process takes place in four stages [12], [13]. A brief description of each stage is given below:
a) Modelling – The robot/machine must first be mathematically modelled in order to relate its functioning to its control system