Analysis and optimization of injection molding process on warpage based on Taguchi design and PSO algorithm
Lei Zhang, Tien-Li Chang, Chung‐Chen Tsao, Kun-Chi Hsieh, Chun‐Yao Hsu
Abstract
Injection molding (IM) is a rapid process that injects molten plastic into a mold cavity for cooling and solidification. It is widely used in daily necessities, auto parts, medical equipment, electronic products, and infant toys. However, the process parameters of IM are complex and interact with each other. IM simulation software can not only provide optimized solutions to process problems, but also improve the quality of manufactured parts. This study report combines Taguchi design and computational software (Moldflow, MINITAB, and MATLAB) to analyze and simulate the process parameters of the back cover of the LCD monitor to obtain its minimum warpage ( W ) in z -direction. The results of Taguchi design (traditional optimization technique) show that the optimal processing parameter combination to obtain W is 40°C of mold temperature (A 1 ), 220°C of plastic temperature (B 1 ), 70 MPa of packing pressure (C 1 ), 5 s of holding time (D 3 ), 20 s of cooling time (E 1 ), 60 cm 3 /s of injection speed (F 1 ), 40% of switch position (G 1 ), and 90 MPa of the injection pressure (H 3 ), named as A 1 B 1 C 1 D 3 E 1 F 1 G 1 H 3 . Four important parameters (B, C, D, and G) were obtained by Taguchi design in IM process. The computed warpage errors, obtained by Moldflow software and regression analysis, are less than 2.5%. However, Taguchi design can only result in an optimal set of combinations of specified processing parameters. The particle swarm optimization (PSO) algorithm, which is a non-traditional optimization technique, often overcomes obstacles faced by Taguchi design. The examined warpage errors through Moldflow software and PSO are less than 3.5%. It is shown that the combination of Taguchi design and PSO algorithm is very suitable for processing optimization of IM production. Its smart manufacturing method can not only reduce production costs and time, but also increase the flexibility of processing parameter selection to a greater extent.