Performance comparison of artificial neural network and expert system in prediction of flow stress

Modeling of various manufacturing processes, including force and power requirements, depends on accurate estimation of a material´s flow stress. The paper presents a mutual comparison between rule-based expert system and artificial neural network in predicting flow stress of a commonly used type of steel. The prediction processes take microstructure, applied temperature, strain, and strain rate as process parameters. The prediction results of both the systems show a good deal of match with the actual values.