Finite element and artificial neural network analysis of ECAP

• Published in: Computational materials science 2012-10, Vol.63, p.127-133
• Main Authors: Esmailzadeh, M., Aghaie-Khafri, M.
• Format: Article
• Language: English
• Subjects: Aluminum alloys; Artificial neural network; Condensed matter: structure, mechanical and thermal properties; Cross-disciplinary physics: materials science; rheology; Deformation and plasticity (including yield, ductility, and superplasticity); Equal channel angular pressing; Exact sciences and technology; Finite element model; Materials science; Mechanical and acoustical properties of condensed matter; Mechanical properties of solids; Other heat and thermomechanical treatments; Physics; Strain measurements; Treatment of materials and its effects on microstructure and properties
• ISSN: 0927-0256; 1879-0801
• DOI: 10.1016/j.commatsci.2012.05.075

► FEM and ANN were used to simulate ECAP deformation of AA2024 aluminum alloy. ► The effect of friction on strain distribution along the sample width was studied. ► FEM and ANN results were compared with experimental measurements. Equal channel angular pressing (ECAP) is the most promising among the developed severe plastic deformation (SPD) techniques to induce strain in bulk metals. In this study finite element method (FEM) and artificial neural network (ANN) were used to simulate ECAP deformation of AA2024 aluminum alloy. The results show that the equivalent plastic strains are not uniform and the deformation inhomogeneity indexes and the location of maximum equivalent plastic strain are varied with the increasing friction coefficient. Moreover, the area over which friction acts and hence the total accumulated friction force is reduced when the billet length is reduced. The FEM and ANN results were in good agreement with experimental measurements.