- 作者: 趙平宜; 黃永東
- 作者服務機構: 中山大學機械工程學系所
- 中文摘要: 本研究之目的在利用田口式實驗設計法(Taguchi's experimental design)規劃金屬切削實驗,以評估刀具壽命及表面粗糙度的方法。文中以傳統車床對S1017C及S1045C材料做乾式及濕式切削,並選取材質、後頃角(rakeangle)、速度、進給和切深等五個控制變數。另外,亦以CNC車床對硬化鋼做乾式切削,選取材料硬度、跨距(overhang)、速度、進給和切深等五個控制變數。本研究以L16直交表規劃實驗,重複三次實驗,並以訊號雜音比(signal to noise,S/N)及平均值表做資料變異數分析(analysis of variance, ANOVA)。最後,驗證ANOVA表是否常態分佈(normal distribution)及等變異數(constant variance),以確認資料分析結果的合理。本研究是經由少量的實驗後,即可判斷出材料之切削性,且準確性不亞於由傳統要因設計( factorial design)所得之結果。所以,此方法在金屬切削實驗之研究,可節省切削時所花之時間及物力。
- 英文摘要: This paper describes the experience that the authors obtained in using Taguchi’s experimental methodto design metal cutting experiments. The tool wear and the surface roughness of the workpiece werethe major concerns in the experiments.The experiments were conducted on two traditional lathes anda CNC turning lathe. The factors included in our experiments were the work material, the rake angle,the velocity, the feed rate, and the depth of cut in using traditional turning lathes as well as the hardnessof the material, the overhang length of the tool holder, the cutting speed, the feed rate, and the depthof cut on the CNC lathe. In this paper, we introduce the L16 orthogonal array for designing experimentsand running randomized experiments with three replications, and establish ANOVA's (analyses of variance)for the signal-to-noise ratio (S/N) and mean of response. Finally, we verify the ANOVAs for normalityand constant variance. This method requires fewer experiments than does the conventional factorial design.By using this method, reliable analysis of the cutting operation can be carried out with less investmentin experiments.
- 中文關鍵字: ANOVA; signal-to-noise; normality; constant variance
- 英文關鍵字: --