| Summary: | 碩士 === 國立屏東科技大學 === 機械工程系所 === 106 === In this study, we use the milling nickel-based Inconel 718 turbine blade of the five-axis machine tool to explore how the optimal combination of parameters can be obtained from the high-speed milling in the course of the curve surface respectively against the single quality characteristic of the surface roughness and the processing time. Due to the inconsistency in the optimal combination of parameters between such two quality characteristics as the surface roughness and the processing time, we further utilize the grey relational analysis as well as the fuzzy inference system to obtain their multiple quality characteristics. First of all, we make use of the Taguchi’s robust manufacturing process design to conduct the experiment by planning the parameter combination in terms of the L16(45) orthogonal arrays, measure each surface roughness posterior to the experiment and record the processing time of the experiment. Based on the experimental results, we analyze the signal to noise ratio (S/N Ratio) of each quality characteristic and separately obtain the optimal parameter combination of single quality characteristic of the surface roughness and the processing time. Next we take the Ra 0.4 μm of the surface roughness as the restrictive requirement and find out the multiple quality characteristic which simultaneously considers the surface roughness and the processing time by way of the grey relational analysis and the fuzzy inference system. As the study results indicate, when the optimal parameter combination of the multiple quality characteristic is adopted in the processing, the mean value of the surface roughness Ra we obtain is 0.3924μm, and took 325 sec, which corresponds to the restrictive conditions. Compared to the Taguchi’s robust manufacturing process experiment, in which the surface roughness Ra meets the requirement of less than 0.4μm with the 564 sec that the shortest processing time, the processing efficiency is improved by 42.3%. Via the grey relational analysis and the fuzzy inference system, this study integrates two different quality characteristics of the surface roughness and the processing time into the optimal parameter combination of one single quality characteristic. The results prove that the processing efficiency can be actually upgraded, concurrently taking the quality objective into consideration. Besides, the users can alter the restrictive conditions, depending on their needs.
|
|---|
