Optimization of Process Parameters during Pressure Die Casting of A380: a Silicon-Based Aluminium Alloy Using GA & Fuzzy Methodology

Abstract

Pressure die-casting is generally useful in casting of zinc, Aluminium, magnesium, lead and tin-based alloys. Better surface finish, tolerances and dimensional accuracy can be achieved by optimization of significant process parameters of Pressure die casting (PDC) such as solidifying-time, melting-temperature, fill-up-time, inoculation-pressure and plunger-velocity etc. In addition, with the proper selection of optimum process parameters the PDC defects such as flash, cold-shut, porosity, misrun, etc. also minimized. In the present work, minimization of the PDC defects using artificial intelligent technique comprises of Genetic-algorithm (GA) and fuzzy-logic-methodology (FLM). A380: A silicon based aluminium alloy carburetor-housing is considered for the same. L18 orthogonal array based Taguchi design approach is considered for number of % defects per batch in a lot of 500 components as FLM knowledge base. To estimate the extent of accuracy of predictions made by FLM, regression analysis isperformed. FLM is also validated with R-square approach. Besides this, ANOVA has also been applied to estimate the significance level in terms of percent contribution among the input parameters and defects during PDC. Solidifying-time & fill-up-time has been found most sensitive parameter. Melting-temperature, plunger-velocity and inoculation-pressure are also identified as influencing parameters with comparatively low percent contribution. To optimize process parameters GA is applied and the fitness value is calculated using FLM. From this combination of artificial intelligent approach an optimum combination of process parameter has been obtained. Further, the results are compared with the Taguchi methodology and the approach provides appreciable amount of decrease in PDC-defects by fine tuning the input parameter.