Numerical and experimental modeling of multiple pass radial forging of Alloy 718

Radial forging of a small diameter Alloy 718 billet was modeled under non-isothermal and axisymmetric conditions using an Updated Lagrangian based finite element code. In this model, each stroke of the radial forging tools was modeled as a separate simulation. An efficient method to account for the non-contact time of the forging tools between deformation cycles was developed in order to improve transient temperature predictions during each pass. An improved representation of the chuckhead boundary condition constraint was also developed. A three pass forging sequence using Alloy 718 billet was conducted on a four die radial forging machine to validate the finite element model. Results show that deformation was uniform along the workpiece length while effective strain varied with radial position. Temperature distributions showed that significant differences develop between the workpiece surface and the interior during transport and are maintained during forging. Methods were developed to measure the temperature profile and effective strain distribution in the billet after forging to validate the FEM models and are presented.; Controlled high temperature compression testing was used to simulate multiple pass radial forging and to develop a predictive relation for as-forged grain size. To simulate the radial forging process, experimental data was generated using a Gleeble thermomechanical testing machine and samples taken from an Alloy 718 billet preform. The independent variables considered were strain per pass, number of forging passes, time per pass, and temperature. Analysis of Variance was employed to test the data for significant factors and interactions. Strain per pass and temperature were found to be the dominant factors in affecting the microstructure, whereas, the number of passes and time per pass were not found to be as effective in controlling the as-forged grain size. The predictive relation which relates grain size to the strain per pass and the starting grain size is based on static recrystallization and was found to provide good fit for the experimental data. The relation is valid for super-solvus forging temperatures.