| Summary: | 碩士 === 國立臺灣師範大學 === 工業教育學系 === 94 === The purpose of this study was to obtain the optimal welding process parameters in experimental range by the taguchi method with the grey relational analysis. The correlated results and discussions were focused on multiple quality characteristics as weld microhardness, weld width, and weld undercut.
In experiments, the gas tungsten arc welding was used. The base metal were the sheets of type 316 stainless steel . The welding method was butt welding with filler by a servo mechanism. In the experimental design, a L18 orthogonal array was used to reduce the number of welding experiments and the selection of process parameters inclueded the arc length, welding current, welding speed, root openings, filler speed and nitrogen content in shelding gas.
In the single quality characteristic, the experimental result showed that the average of weld microhardness was Hv194.2 for the optimal process parameter and Hv178.8 for the initial process parameter. The hardness was arised about 9%. The average of weld width was 7.88mm for the optimal process parameter and 10.15mm for the initial process parameter. The width was decreased in 2.27mm. The weld of the optimal process parameter indicated no undercut. The weld of the initial process parameter indicated serious undercut.
In the optimal process parameter of multiple quality characteristics, the experimental result showed that the average of weld microhardness was Hv178.5, the average of weld width was 8.23mm and the weld was no undercut. All of the optimal process parameters had good reproducibilities which were less than 30%.
When the shilding gas was pure argon, the microstructure of weld showed ferritic-austenitic morphology. When the nitrogen content in the shelding gas was 1%, the microstructure of weld showed ferritic-austenitic morphology but the ferrites were semicontinous. When the nitrogen content in the shelding gas was 2%, the microstructure of weld showed the mixture of ferritic-austenitic morph-
ology and austenitic-ferritic morphology.
In the further, there is an following advice of reserch : It is possible to design a curve by applying dynamic characteristics in the taguchi method. To do this, we can reach the goal of optimal process welding parameter just change signal factros. It is worth to make efforts on this probem.
|
|---|
