標題: 田口方法於馬口鐵蓋切線生□改善之應用The Application of Taguchi Method to Improve the Score Line Corrosion for Full Aperture tinplate Easy-Open-Ends 作者: 彭雲鄉Yuan-Hsiang Peng蘇朝墩Dr. Chao-Ton Su管理學院工業工程與管理學程 關鍵字: 參數設計;直交表;信號雜音比;累積分析法;田口方法;亞米加轉換;Parameter Design;Orthogonal array;Signal to noise;Attribute accumulation analysis;Taguchi method;Omega transformation 公開日期: 2000 摘要: 摘要 易開蓋式馬口鐵蓋為使開啟順利，易開蓋必需在蓋面預先切下一道切線(score line)，當用拉環拉起時，蓋面即沿此切線順行開啟。然而，在切線加工過程中，已破壞蓋面之塗漆保護層，雖有再重噴塗一道塗料在切線上，但還是常可發現切線生□情形，不但外觀不雅，而且破壞了魚罐頭製造廠的品質形象，甚至偶而還會遭致客戶索賠。 本論文應用田口玄一(Genichi Taguchi)博士所發展的參數設計(parameter design)方法，找出馬口鐵蓋切線塗漆重噴塗最佳化的操作條件設定參數，使馬口鐵蓋切線重噴塗能完全，而不會在放置極短的時間即發生生□現象。切線重噴塗製程改善的品質特性，經廠內技術部門多位工程師與品保人員研討結果，決定以切線部位經硫酸銅浸泡後，出現腐蝕黑點數為單一品質特性，配合直交表(orthogonal arrays)的配置，藉以加速實驗的進行並簡化分析。 本論文以望小特性(Smaller-The-Better)信號雜音比 (signal-to-noise ratio，SN比)和等級分類型資料(ordered categorical data)同時進行分析，研究結果顯示，這兩種分析方法獲得相同的最佳參數設定，經最後進行確認實驗後，發現加法模型在本實驗中是合宜的，並改善了切線生□的情形。Abstract In order to facilitate the consumers to easily open a tin-plate steel Easy-open-end, the end manufacturer made a score line on the end panel. While we open the end by pulling tab, score line break and breakage progressing along this score line to form an open panel. In fact, the coating applied to protect tin-plate have been destroyed during the score line process. Although a repaired coating was applied after score process, a score line corrosion was discovered occasionally. Such rusting score line not only impacts appearance of cans, also damaged the quality image of fish cans manufacturer. Moreover, customers raised claim on this defect sometimes. The firm applied Taguchi’s approach to the score line process, so as to prevent the corrosion from occurring in short time. A greater technical understanding of the problem by means of brainstorming from engineers and quality assurance staff, then the characteristic was decided by addressing the real problem of score line repairing coating process. The result of score line after dipping by cupric sulfate liquid with single characteristic, namely the number of black spot after was discussed. The study was also analyzed along with appropriate orthogonal arrays to accelerate and simplify the process of experiment. This study based on Smaller-The-Better, Signal-to-Noise ratio, SN ratio and ordered categorical data to analyze. The results show that two methods result in the same optimal parameter design, furthermore the additive model is appropriate in the experiment by means of confirmation experiment. The conclusion do improve the corrosion of score line problem. URI: http://140.113.39.130/cdrfb3/record/nctu/#NT891031003http://hdl.handle.net/11536/67984 顯示於類別： 畢業論文