Flexibility, Automation in Small Metalcasting Facilities
Two case studies illustrate how automation can be implemented in a small metalcasting facility in order to react to today’s global manufacturing demands for quickly delivered customized parts.
Rhythm Wadhwa, Norwegian University of Scence & Technology-Valgrinda, Department of Production & Quality Engineering, Trondheim, Norway
(Click here to see the story as it appears in the October issue of Modern Casting.)
In recent years, flexibility has attracted significant attention from small to medium enterprises (SMEs) in metalcasting and academia due to varying customer demands and increasing competition. Changing operating conditions are forcing firms to be flexible in handling variations in demand and product and uncertainty and changes in the environment. Such factors have affected manufacturing companies for a long time, but their influence has escalated during the past 20 years as a result of advances in manufacturing technology and demand for mass customization.
Organizations, both large and small, require reconfigurable equipment to produce one-of-a-kind or small batch quantities of customized products. Client demand for small volumes of customizable product leads to a paradigm shift in how effectively an SME, with 10-25 employees on average, would operate to quickly and effectively deliver parts. Identifying best practices is a tricky process that is difficult to implement, which is more noticeable when the companies are SMEs. Typically, SMEs have limited resources and knowledge of automation methodologies. To address this, the Norwegian Research Council started a project to test the concept of a shared flexible manufacturing environment tailored to metalcasting SME requirements.
The participating metalcasting consortium in the project agreed to test part handling automation solutions at two metalcasting facilities in a living lab setting. Living labs began to emerge in early 2000, and the concept has since grown. A precondition in living lab activities is that they are used in a real-world context. During this living lab process, constant feedback for improvement was collected and transformed into a requirement list for the technology providers.
Case Study One: Flexible Robot Part Handling in an Iron Casting Facility
Preferring to remain anonymous, the iron metalcaster is a manufacturer of residential wood stoves. The company faces seasonal customer demand patterns and produces eight series of wood stoves, each with four variants. Two of the series have special coating processes after the individual assembly parts are cast, with special handling requirements. The metalcaster implemented a test automation cell between the cleaning and sorting processes as a potential test area for implementing a flexible automation solution for handling the varying part families. The company wanted to test an automation cell built around the range of products that they manufacture and choose the best way to integrate it into production. The company incorporated a used robot to implement cost effective, quick changeovers for fast, flexible part routing and handling.
The overall design concept of the material handling’s Ethernet communication handling system consisted of three subsystems: the robot manipulator, the material handling system and the vision system. In the cell, an overhead camera identified the orientation of the part lying on the conveyor belt, which was internally tracked by the robot (Fig. 1). The image captured by the camera was processed by the vision system and transferred via closed network connection to the robot. The robot gripper then moved the electromagnets accordingly to pick up the part.
Vision System Challenge: Picking from Conveyor vs. Bin
The vision module was to perform object recognition of a three-dimensional part and locate the coordinates. Markers were cast-in to aid in an accurate estimation of the part’s orientation for a pick-and-place operation from a conveyor belt and stationary bin. The stove components were flat with dimensions varying between 0.3 and 0.9 sq. in. (200 and 600 sq. mm) and thicknesses between 0.2 and 0.3 in. (5 and 8 mm). Picking from a conveyor belt when parts were lying separately at a distance was conveniently implemented for automation. But when parts were combined in random order in a bin, issues arose in identifying the part orientation. Two marking options were tested for this purpose: concentric circles and straight lines (Fig. 2).
The straight line markers were discovered to have a reflection limitation when the tilt angle of the vision system was greater than 30 degrees for iron castings. This made the concentric circles a better option for tracking because a part of the circle was identifiable even under the light of the vision cameras (reflection issues).
Handling the parts once they arrived at the machine was also important. Implementing machine flexibility with the use of robots with reconfigurable grippers and intelligent interfaces (flexible workspace, vision system etc.), automatic tool changers and multi-axis robots helped to enhance material handling flexibility at the most affordable price. Figures 3 and 4 show two gripper prototypes used in the iron facility. The yellow lines indicate the direction of travel of the electromagnet heads to enable part handling encompassing different workspaces.
Case Study Two: Modular Concept for Part Handling & Fixture Flexibility in an Aluminum Casting Facility
In the second case study, an aluminum casting facility wanted to analyze how and where it could use modular components for handling automation. It wanted to look into pallets and overhead conveyors for the internal transport of castings, which would require significant investment and maintenance costs because they would be specially tailored for each new product. During production, parts were transported around the facility on pallets that were linked to different assembly lines. For robotic grippers to handle the parts, they would have to be positioned securely.
Moreover, handling sand cores required additional constraints to ensure they were not worn or cracked during transport. The fixture blocks were adapted to the sand core shapes to ensure enough rest area was available and they would be fixed properly on the pallet.
In the tested automated pick-and-place cell, various manual and motor-propelled jigs and fixtures rotated parts and assemblies through several axes and adjusted to the needs of numerous assemblers. Flexibility in the pallet fixture determined the degree of freedom available to part loading schedules. This flexibility also had to be taken into account when dealing with the ability to accommodate different parts of different shapes and sizes.
As part of the Autocast Project, the aluminum casting facility implemented four modular solutions for handling aluminum parts and sand cores. One of those solutions is shown in Figs. 5 and 6. The same standard base plate was used to transport sand cores and aluminum parts. The aluminum parts were supported by three flexible points. Three arms were adjustable in X, Y and Z directions to ensure all parts within a product family (i.e., swing arms) can be added to the pallet. The implementation presents an integrated solution in which no part must be replaced when the pallet is converted. Aluminum part location pins were adjustable in the height (Z) and longitudinal direction (X). The arms also could be turned around the Z axis of the tubes and rotated independently. When arms were properly adjusted in all directions, they had to be fixed in this position.