On the Formability of Some Metastable Alloy Phases

Pattern recognition and neural network methods have been used to investigate the formability of metastable alloy phases, It has been found that some chemi cal bond parameters Such as valence electron number, electronegativity and metallic radii of cor-nponent elements are the dominating fac tors affecting metastable alloy phase formation. Some semi-empirical rules found in this way may be useful for the construction of expert system for materials design.