| Summary: | 碩士 === 明新科技大學 === 精密機電工程研究所 === 96 === Injection Molding is one of the molding processes, and a popular technique that commonly is used in the industry. It is a complicate manufacturing process; therefore, the product quality is determined by various parameter factors; meanwhile, it may impact its quality characteristics.
This study mainly focused on the fabrication of interior ornaments for vehicles. In this study, Polypropylene (PP) compound material has been selected, and a L16 orthogonal array of Taguchi method was applied to conduct the experiment. Experiments of sixteen experimental runs were based on a Taguchi L16 (29) orthogonal array table. The specimens were prepared under different injection molding conditions by varying the melting temperature, the injection speed, and the injection pressure with two levels for each controlled parameter via three computer-controlled progressive strokes, Furthermore, the average warpage (δ), the tensile stress (σ) and the wear (△m) were three selected quality targets; the average warpage and the wear quality characteristic were Smaller-the-Better; the tensile stress quality characteristic were Larger-the-Better. Through finding the signal to noise ratio, Taguchi method can successfully find the optimal design parameters. In addition, analysis of variance (ANOVA) was applied to identify the importance of parameters to the selected quality targets. Then, a trained Neural Network was used to find the optimal parameter combination and a comparison with Design of Experiments (DOE) was also provided.
The research results showed that the parameter process that obtained from the neural trained process was better than those from DOE. In addition, the neural network used the concept of black-box algorithm that can successfully forecast the average deformation, tensile stress and wear loss of the produced components during the process; thus, the research method is more effective than common approaches.
|
|---|
