Optimization of a Closed Die Forging Process to Manufacture a Gear Wheel by the Use of a Response Surface Model

Heavy-duty components used in the automotive industry, in wind turbines and in many other industrial applications are often produced using hot forging processes. Nowadays the design of hot forging processes aims for the optimization of process efficiency on the one hand and final mechanical product properties on the other hand. Excellent mechanical properties needed for hot-forged components e.g. high load capacity and high fatigue resistance depend on a fine homogeneous microstructure distribution across the final product’s cross-section. Efficiency in hot forging can be optimized by increasing the temperature during processing, which allows for lower forging loads and lower die stresses, thus improving die life in terms of mechanical fatigue. To guarantee for a fine homogenous microstructure across the cross section of the forged good, dynamic recrystallization (DRX) has to be initiated during deformation and Grain Growth (GG) has to be avoided during dwell times and cooling. Due to the high computational costs of finite element simulations an optimization aiming for lowest possible forging loads and finest possible grain sizes is very time-consuming. In this paper a Response Surface Model (RSM) of the forging process is introduced, which allows for much faster evaluation of the outcome of forging simulations, albeit by interpolation of simulation results, and thus allows for optimization. The information required to create the RSM is obtained by Design Of Experiments (DOE) techniques using an FE-model of the forging process which was calibrated earlier. The process variables considered include the initial temperature of the billet and the die kinematics. Subsequently, an optimization algorithm is combined with the RSM to find the design variables giving minimum possible loads during deformation and finest possible grain sizes in the forged product. The RSMs results are validated by the use of the existing FE-model.