| Summary: | 碩士 === 明新科技大學 === 精密機電工程研究所 === 100 === Abstract
This paper applies Design-Expert to generate the technology of D-optimal mixture design which integrating response surface methodology(RSM), and back-propagation neural network (BPNN) integrate the simulated annealing algorithm (SAA) method separately to discuss variation of the thermal conductivity (TC) and coefficient of thermal expansion (CTE) and flexure strength depended on injection molding mixture ratio of 10-20% short glass fiber (SGF) 4-12% polytetrafluoroethylene (PTFE) and 68-86% reinforced polycarbonate (PC) composites.
The analysis of variance (ANOVA) was applied to identify the effect of mixture ratio of SGF and PTFE reinforced PC composites for the thermal conductivity and thermal expansion coefficient and flexure strength. By regression analysis, a mathematical predictive model of the thermal conductivity, coefficient of thermal expansion and flexure strength were developed in terms of the mixture ratio setting. The combining BPNN/SAA optimization method can be obtained for the appropriate combinations of the optimal mixture ratio setting. In addition, the result of BPNN integrating SAA was also predictive with BPNN approach.
The results show that the optimal mixture ratio setting gives appropriate combinations with a PC of 0.79, a SGF of 0.17, and a PTFE of 0.04 by RSM approach. Additionally, BPNN/SAA approaches are gives appropriate combinations with a PC of 0.80, a SGF of 0.16, and a PTFE of 0.04. By verification results show the proposed algorithm of SAA approach has better prediction result than the RSM method.
Keywords: Mixture design, RSM, Thermal Properties, Back-Propagation Neural Network, Simulate Anneal Arithmetic.
|
|---|
