Dependence of martensite start temperature on fine austenite grain size

It has been broadly reported that determination of the martensite start temperature in steels, M-s, requires a complete description of their chemical composition. Recently, several neural networks models considering both chemical composition and austenite grain size (AGS) have been developed. Such models predict a moderate dependence of M-s with AGS. The present work examines the validity of existing neural network models, but focusing on fine AGS (below 5 mu m). (c) 2007 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.