Comparison of analytical methods and AI tools for material characterisation in hot forming

Hot forming processes probably represent the most ancient of forming operations and what is more they are still today commonly used in modern mechanical industry in order to obtain sound parts, achieving large deformations with a limited required power. Hot metal forming operations are characterised by a large number of physical and thermal phenomena which have to be taken into account in order to model and design the processes themselves. Actually several thermally activated phenomena occur during the forming processes such as recovery, recrystallisation, grain growth, precipitation, allotropic transformations, etc. In this paper the comparison between an analytical method based on the Gauss–Newton algorithm and the genetic algorithms (GAs) is proposed with the aim of characterising material behaviour in hot forming operations. Such approaches were utilised in order to determine the coefficients of one of the most effective equation utilised for material characterisation, namely the equation proposed by Beynon.