Optimization of injection molding parameters by data mining method in PIM process

Abstract

Data Mining is a method that can be used to analyze large amount of data and produce useful information. In this study, clustering which is one of data mining tasks is used to clustered machine based on the injection moulding data. This paper is the first documented results on the optimization of Injection Moulding via Data Mining. Powder injection moulding is a process to produce near net shape with intricate part in mass production. This work focus on the optimization of injection molding process with combination of fine, coarse and bimodal water atomized SS 316L powder particles. The parameters involved in the optimization are injection pressure, injection temperature, mould temperature, holding pressure, injection rate, holding time, powder loading, cooling time and particle size. These variables are based on the defect score, green density and green strength. The key influencer report shows that the most influencing factors are injection rate, holding pressure as well as mould temperature where defect score lower than 2.4 can be achieved. The density higher than 5.34g/cm3 is also influenced most by the mould temperature. The result also shows that the optimize condition can be achieved by using bimodal particle. Injection rate and mould temperature gives the highest impact on the defect score and green strength value. While highest green density is significantly affected by powder loading and injection pressure.