Finite element modelling approaches to the accurate dimensional prediction for a cold-forged part

Abstract

The dimensions of the cold-forged part turn out to be larger than those of the die cavity, since a die can be deformed more elastically than a forged material. To predict precisely the dimensions of the forged part and to determine the die dimensions for net-shape components, dimensional evolution of the die and workpiece should be performed carefully; therefore, finite element methods (FEMs) have been proposed to predict more accurately the dimensions of the forged product in closed-die upsetting. Together with finite element analysis (FEA), direct measurements of strains from the die during forging were attempted in order to validate the simulated results. In the present study, the dimensional changes of the die and workpiece due to elastic and thermal effects were investigated thoroughly by separating a single forging step into three stages: loading, unloading and ejecting stages. As a result, FEM results could predict the part dimensions to within a range of 10 μm. In particular, the characteristics of the unloading and ejecting stages have been discussed in conjunction with the elastic recovery of forged parts. When temperature changes due to the deformation heat were additionally considered in the finite element models, the predicted dimensions more closely coincided with the experimental results. Since the elastic modulus of workpiece materials has a great influence on the final forged product, the effects of the elastic modulus for the workpiece were deliberately investigated, and then the modelling technique applicable to the three-dimensional forged part was proposed in order to increase the simulation efficiency. The dimensional differences between the die and forged part were compared quantitatively in the experimental and FEA. The result was used for machining a modified die. The FEM results using the modified die validated that the dimensional difference between the die and forged part had a linear relationship.