Computational and Experimental Adjustment of the Databases for Computer Simulation of Cast Technologies

Abstract

Results of the full-scale experiments on crystallization of test casting of 20GL steel used in manufacturing cast parts of the freight car bogies are presented. The data of the temperature measurements (thermocouples located within the mold) are used to solve the inverse problem of recovering thermophysical characteristics of mold materials in the heat conduction equation. The proposed calculation-experimental recovery procedure can be used for data replenishment of the material databases of specialized CAE software (Computer Aided Engineering) intended for simulation of foundry technologies. The method consists in a series of computational experiments, identical to full-scale ones thus bringing the calculated temperature curves close to the full-scale readings of thermocouples due to directional changes in the dynamic variables using Levenberg-Marquardt multivariate optimization.