Application of Bayesian Analysis Method in the Design Optimization of Permanent Casting Mold

This work is a case study of applying Bayesian analysis, a statistical data method, in the design optimization of permanent test-bar mold. The permanent test-bar mold is used in casting foundry to examine the metal quality. Since the current standard test-bar mold suffers from shrinkage porosity which detracts from best properties, a modified design is recently proposed to improve the mechanical properties. In order to validate the new design, Bayesian data analysis method is utilized to analyze the experimental data from the two designs. The effects of the mold designs and casting process operational parameters on the mechanical properties of castings are compared. Main effect to the mechanical properties is identified based on the Bayesian analysis.