Study on microstructural evolution of near-net shape forging process of AA6082 wheel – Dynamic recrystallization and processing map analysis

Abstract

The hot deformation behavior of AA6082 was investigated through isothermal compression tests performed within a temperature range of 350 °C to 560 °C and a strain rate range of 0.05 to 15 s⁻¹ to replicate the actual process of a wheel near-net shape forging process using finite element (FE)-based numerical simulations. The near-net shape forging technology of an AA6082 commercial wheel consists of three stages: preforming, rough forging, and finish forging. The constitutive model of flow stress data was numerically integrated using a novel strategy that incorporates nonlinear regression methods on the mathematical models of up-to-peak, dynamic recovery (DRV), and dynamic recrystallization (DRX) behaviors of the material during isothermal compression tests. A processing map analysis was established based on the flow stress data, and the obtained parameters, such as power dissipation efficiency and instability, were integrated into the numerical model using trilinear interpolation with a clamping method. The primary contribution of this study is to evaluate the effectiveness of the numerically integrated processing map parameters in predicting microstructural grain evolutions during the wheel near-net shape forging process. Microstructures of the material in several areas of the workpiece after each forging stage were examined using an optical microscope. Subsequently, the Johnson-Mehl-Avrami-Kolmogorov (JMAK) numerical model of dynamic grain size was derived from a combination of experimental observations and numerical processing parameters to establish a prediction of average grain size in all areas of the workpieces during each stage. Overall results demonstrate that the new strategy employed in constitutive modeling is practical, efficient, and highly accurate in modeling the high-noise flow stress curves of AA6082 material, achieving an average absolute relative error of 1.88 %. The new numerical model for average grain size prediction closely aligns with experimental measurements, exhibiting errors below 6 % in most observation areas. However, one area was identified with a 12.88 % error, attributed to a grain growth defect. The new numerical simulation results of processing map parameters are promising in predicting recrystallization conditions and accurately identifying defect-prone areas. Ultimately, the near-net shape forging technology successfully refined the grain size on the wheel by 53.42 % to 61.66 %, and the uniformity of grain size increased to 92.43 % and 95.46 % after the rough and finish forging processes, respectively.