engleski

Ductile iron microstructure prediction based on melt thermal analysis using artificial neural networks

The paper presents the results of application of artificial neural networks in predicting the ductile iron graphite microstructure, i.e. nodularity and nodule count. The input parameters are the data acquired through thermal analysis of the melt, namely the ones that affect nodule formation the most: liquidus temperature, eutectic undercooling temperature, recalescence, solidus temperature, graphite factor 1 and graphite factor 2, near-solidus cooling rate, and eutectoid temperature. Good accordance between measured and predicted data is observed so the proposed neural network method can be successfully applied for predicting the graphite formation in ductile iron based on melt thermal analysis.

artificial neural network ; ductile cast iron ; thermal analysis ; graphite

nije evidentirano