This was a collaborative project the American Foundry Society, DOE and Idaho State University to investigate the application of advanced intelligent control methods to the cupola melter, a furnace used for melting pig iron, scrap steel, cast iron scrap, foundry return scrap, and ferroalloys to a specified tapping temperature and chemical composition. The foundry cupola has historically been the primary method for melting iron because of its low cost and simplicity. Recently, however, the need for pollution control devices and foreign competition has contributed to a decline in the domestic market. One way to help improve the role of the cupola in the foundry industry is to introduce automatic control technology to lower material and processing costs and improve product quality.
The Cupola Neural Network Model, now available from the American Foundry Society, is allowing foundry operators to increase control over, and reduce energy use in, cupola furnaces