A model for evolution of as-cast microstructure in ductile iron alloy

This thesis concerns the application of micro-modeling of ductile iron solidification to development of a process control methodology for simulation of microstructure evolution and prediction of shrinkage tendency. The approach is to utilize a simple lumped system heat transfer model which incorporates a new formulation for simulating eutectic ductile iron solidification to analyze and interpret cooling curve data routinely acquired for process control purposes. A key feature of the eutectic solidification micro-model is that the graphite kinetics is de-coupled from the austenite kinetics which are completely determined by thermodynamic and mass balance considerations. In addition, the graphite kinetics expression accounts for the evolution of a lognormally distributed nodule size distribution. An empirical two-parameter continuous nucleation model is proposed. Solute redistribution in the ternary Fe-C-Si system is considered. The influence of magnesium on undercooling the melt is simulated and studied using experimental data. A semi-empirical model has been proposed to simulate the eutectoid reaction and predict the as-cast microstructure in ductile iron. A novel method is used to evaluate the kinetics coefficients in the simulation models. Instead of the analysis of grain growth kinetics, an optimization algorithm is utilized to determine values of the 'calibration' parameters in the graphite kinetics and the solid state kinetics models on the basis of minimization of the sum of squares of deviation between simulated and experimental cooling curves and criteria related to microstructure prediction. The micro-model has been used to analyze a comprehensive data set of cooling curves obtained in laboratory experiments to determine the dependence and sensitivity of the 'calibration' parameters to processing conditions.