FINAL REPORT: Reduction in Energy Consumption and Variability in Steel Foundry Operations

This project worked to improve the efficiency of the steel casting industry by reducing the variability that occurs because of process and product variation. The project focused on the post shakeout operations since roughly half of the production costs are in this area. These improvements will reduce the amount of variability, making it easier to manage the operation and improve the competitiveness. The reduction in variability will also reduce the need for many rework operations, which will result in a direct reduction of energy usage, particularly by the reduction of repeated heat treatment operations. Further energy savings will be realized from the reduction of scrap and reduced handling. Field studies were conducted at ten steel foundries that represented the U.S. steel casting industry, for a total of over 100 weeks of production observation. These studies quantified the amount of variability, and looked toward determining the source. A focus of the data collected was the grinding operations since this is a major effort in the cleaning room, and it represents the overall casting quality. The grinding was divided into two categories, expected and unexpected. Expected grinding is that in which the location of the effort is known prior to making the casting, such as smoothing parting lines, gates, and riser contacts. Unexpected grinding, which was approximately 80% of the effort, was done to improve the surfaces at weld repair locations, to rectify burnt on sand, and other surface anomalies at random locations. Unexpected grinding represents about 80% of the grinding effort. By quantifying this effort, the project raised awareness within the industry and the industry is continuing to make improvements. The field studies showed that the amount of variation of grinding operations (normalized because of the diverse set of parts studied) was very consistent across the industry. The field studies identified several specific sources that individually contributed to large process variation. This indicates the need for ongoing monitoring of the process and system to quantify the effort being expended. A system to measure the grinding effort was investigated but did not prove to be successful. A weld wire counting system was shown to be very successful in tracking casting quality by monitoring the quantity of weld wire being expended on a per casting basis. Further use of such systems is highly recommended. The field studies showed that the visual inspection process for the casting surface was a potentially large source of process variation. Measurement system analysis studies were conducted at three steel casting producers. The tests measured the consistency of the inspectors in identifying the same surface anomalies. The repeatability (variation of the same operator inspecting the same casting) was found to be relatively consistent across the companies at about 60-70%. However, this is still are very large amount of variation. Reproducibility (variation of different operators inspecting the same casting) was worse, ranging between 20 to 80% at the three locations. This large amount of variation shows that there is a great opportunity for improvement. Falsely identifying anomalies for reworking will cause increased expense and energy consumption. This is particularly true if a weld repair and repeated heat treatment is required. However, not identifying an anomaly could also result in future rework processing, a customer return, or scrap. To help alleviate this problem, casting surface comparator plates were developed and distributed to the industry. These plates are very inexpensive which enables them to be provided to all those involved with casting surface quality, such as operators, inspectors, sales, and management