This thesis work has been performed at the Division of Assembly Technology at Linköping University for a period of ten weeks. The work is part of the project ProFlexA, which is a research project in collaboration between Linköping University, Lund University, and industrial partners such as Combi Wear Parts and Smålands Stålgjuteri AB.
The casting process creates random incurred burrs and defects, which must be removed during a cleaning process to make sure that the cast can be sold. Automation has long been sought for the cleaning of castings due to, among other things, poor working conditions for staff, but this has proved difficult to implement with today's technology. Automation has for some time been available for more capital strong companies with large annual volumes, but for smaller companies with small-and medium-sized volumes, automation has not been an option.
The goal of ProFlexA is to develop an automated cleaning cell aimed at small-and medium-sized volumes.
The goal of the thesis is to do a pre study at ProFlexA’s partners on cleaning of castings, and develop concepts for automated cleaning of castings.
By recording the process and cycle times, the cleaning of castings has been documented as it works today, and on this basis, information and statistics has been compiled. Conclusions have been drawn on preventive measures that may facilitate the development of an automated cleaning cell. One of these is to change the geometry of the steel riser so that the resulting stub after removal by the pneumatic hammer is more controlled, and another measure is to relocate the location of the riser stub to an equal side of all the products. It is also important to prevent the defect caused by sand residues.
Based on the documentation, it has been revealed that the product often is over-worked, which implies that an automated cell would provide a more even quality.
Forces that occur in an operation of cleaning of castings have been measured. The results suggest that the forces in this context are 100 to 180 Newton, which indicates that the robot selected to the project, IRB 6660, is likely to be oversized. It can therefore be discussed if it is relevant with the current safety factor or if there is a reason to search for a smaller robot.
Proposals have been prepared on concepts of automated cleaning of castings. These concepts show how a robotic cell can work when a single robot, compared with two robots are part of a cell, but also how it works with stationary tools compared with mobile tools.
The design of a robotic cell should be sought in the objectives and needs of the specific company. A robotic cell with two robots is for example to be preferred when the need for efficiency is higher. Tools to a robotic cell is built in the simplest possible terms by a motor, a spindle and a tool, which is the option that seems to be a good choice in terms of whole and investment friendliness.
For Combi Wear Parts a robotic cell with two robots is suitable, and for Smålands Stålgjuteri AB a robotic cell with a single robot is probably the best choice. For both of these solutions small and economic stationary tools would probably suit well.