This presentation was made at the 2019 NAFEMS World Congress in Quebec Canada

Resource Abstract

BAIMTECH and BIAM are the leading manufacturer of aircraft engine castings in China. In 2015, BAIMTECH and BIAM successfully implemented a Casting Expert System with a focus on real casting processes comprising design, production and testing (physical and virtual).



This Casting Expert System collects and manages the historical data generated by the actual (physical) casting processes as well as material and simulation data into a data management system. In a first phase the implementation focused on data gathering, managing and structuring the data for analysis. The collected historical data were collated and summarized to describe the whole casting process and to serve as input parameter for digital casting engineering based on simulation and allowing to defect prediction. The Casting Expert System combines various methods, technologies and processes in manufacturing casting. It established the casting process database containing material parameters, typical castings, pouring systems, process parameter, equipment, tooling module and other relevant content. Casting wizard templates take multiple processes into consideration, e.g. Requirement Formulation, Process Design, Modelling, Melting, Pouring, Measure, X-Ray, Heat Treatment, Fluorescence and Mechanical Testing. Based on an integration and implementation of processes product development cycle shortened, the number of human errors in casting manufacturing decreased and notable cost savings were achieved.



In later implementation phases integrated standard casting simulation process templates provide virtual correlation analysis and guidance for designing the casting process. Later on, new methods related to machine learning and data mining were introduced to the Casting Expert System. Based on the use and interplay of machine learning methods, data mining and CAE tools, the Casting Expert System provides engineers a solution to detect defect factors and to apply a parameter importance analysis. Detection of defects based on single parameters enables optimization of process parameter and leads to improved product quality prediction and finally allows casting process recommendations.