Foundry Data Analytics to Identify Critical Parameters Affecting Mechanical Properties of Investment Castings

Abstract

Besides shape fidelity and internal soundness, mechanical properties have become critical acceptance criteria for investment cast parts. These properties are mainly driven by the chemical composition of cast alloy as well as process parameters. It is however, difficult to identify the most critical parameters and their specific values influencing the mechanical properties. This is achieved in the present work by employing foundry data analytic based on Bayesian inference to compute the values of posterior probability for each input parameter. This is demonstrated on real-life data collected from an industrial foundry. Controlling the identified parameters within the specific range of values resulted in improved mechanical properties. Unlike computer simulation, artificial neural networks and statistical methods explored by earlier researchers, the proposed approach is easy to implement in industry for controlling and optimizing the parameters to achieve the desired range of mechanical properties. The current work also shows the way forward for building similar systems for other casting and manufacturing processes.