Quality improvement in investment castings using genetic algorithm

Abstract

Investment casting is well-known for its distinguished characteristics such as manufacturing small industrial components of ferrous as well as nonferrous alloys used in aerospace, automobile, bio-medical, chemical, defense, etc. with closed tolerances at relatively low cost. These industrial components need to be defect free as well as must possess desired mechanical properties. This quality metrics (defect free castings with desired mechanical properties) is mainly driven by process parameters associated with different sub-processes of investment casting including wax pattern making, shell making, dewaxing, melting & pouring, and chemical composition of alloys. It is always challenging to identify such parameters affecting quality of investment castings. In this work, an application of Genetic Algorithm has been extended to identify critical parameters and their specific set of values affecting quality of investment castings. This technique is found be very useful in performing data analytics.