Deconvolution of superimposing reaction signals from DSC curves in selected Al-Mg-Si-(Cu) alloys by mean-field modeling and HEXRD

The present work describes a systematic approach to improve mean-field simulations of dissolution and precipitation of precipitates from the late stages in the precipitation sequence in AlMgSiCu alloys for processing simulations. In addition to metastable precipitates, it considers the evolution of two populations (coarse and fine) of the thermodynamically stable phases in EN AW-6061 and EN AW-6016, β-Mg₂Si and Si respectively. The setup is based on two previous publications using heterogeneous nucleation site energies and their distribution. The simulations are calibrated using data from continuous cooling and heating experiments obtained with differential scanning calorimetry (DSC) and high-energy X-ray diffraction (HEXRD). We show that mean-field simulations combined with DSC and HEXRD can provide valuable information to eliminate uncertainties related to, for instance, the thermodynamic description of metastable precipitates. The method described in this paper also delivers very reasonable results for the evolution of the individual phase fractions. It potentially facilitates the assessment of the influence of different types of nucleation sites and their densities.