Optimization of Process Parameters Using Taguchi ‎and ANOVA

The use of the best operating parameters is needed to produce better quality of plastic products. Since the quality of products is mostly influenced by process condition, determining the optimum process condition with a few experiments is a key task. Melting temperature, injection pressure and cooling time with three levels each were selected which directly affect the dimensional shrinkages in injection molding process of beverage crate product. Shrinkage is chosen as a response variable since it is cause of many defects. Taguchi method was selected for the optimization of these parameters. The objective of this research is to reduce the shrinkage of a crate. Nine types of crates were produced using Taguchi Design of experiment (L۹) approach. Minitab statistical software package were used to analyze experimental results. The shrinkage defects are the “smaller the better S/N ratio” type of quality characteristics. Significant parameters were identified through ANOVA at ۹۵% confidence level. Optimal values for melting temperature, injection pressure and cooling time were found ۲۹۰ C, ۱۳۰ MPa, and ۶۰ seconds, respectively. From analysis of variance, melting temperature was the most significant factor (۱۴.۷ %), cooling time was the second (۱۱.۵ %) and injection pressure was the least significant (۹ %). The predicted optimum response variable (shrinkage) was found as ۶.۳۰ mm and the conformation experimental test gave ۶.۱۱mm. The margin error of ۳% supported the acceptance of the confirmation test result of ۶.۱۱ mm. The shrinkage is reduced from ۶.۸۱ mm to ۶.۱۱ mm which is a ۱۰.۳% reduction. The identification of the influence of parameters is believed as a key factor in assisting injection molding process designers in determining optimum process conditions. Therefore, the implication of this research is that robust optimization approach withstands the injection molding process variations in a more realistic way.

کلیدواژه ها:

Taguchi design of experiment ، ‎Shrinkage ، Injection Molding ، ‎ANOVA ، S/N.‎