A mechanism-based model for prediction of damage and failure in cold forging processes
Cold-forged steel parts are widely used in the automotive and mechanical engineering industry due to the increasing demands for quality and reliability. The evolution of material damage and failure in complex forming processes still cannot be accurately predicted by means of the available finite element (FE) codes. One of the main reasons is that the real mechanisms of ductile damage in metals which include nucleation, growth, and coalescence of voids are not considered in industrial practice. Simplified phenomenological approaches are often used in which a macromechanical damage criterion is formulated depending usually only on the current stress-strain state. The main goal of the present study was to improve a computer-aided design of cold forging processes through advanced material modeling in the FE simulations of damage and failure using physically based material models. Detailed experimental investigations revealed that brittle non-metallic inclusions (e.g. manganese sulfides) act as initiators of material degradation by voids which nucleate due to the particle cracking or decohesion of the surrounding steel matrix depending on loading conditions. Based on these microstructural observations, the well-known Gurson-Tvergaard-Needleman (GTN) material model [5, 9] was extended to account for the damage mechanisms relevant to cold forging of steels. The developed model was implemented as user subroutines UMAT and VUMAT in the commercial FE code ABAQUS and was appropriately validated. Simulation of industrial cold forging processes showed that the extended GTN model can often better predict the location of damage as compared to the standard GTN model and other damage criteria commonly used in practice such as the Cockcroft-Latham criterion . After a proper fit of the model parameters using conventional tensile and compression tests, the estimation of the moment of material failure in multistage forging processes provided satisfactorily accurate results.