Resumen
Experimental evaluations based on statistical and scientific criteria play fundamental role in identification of optimal performance parameters in manufacturing engineering. Based on this premise, the Doehlert?s experimental planning method was applied to external cylindrical turning process of ABNT 1045 steel, with the purpose of determining the roughness correlation as a function of the cutting speed (Vc) and feed rate (f). The Doehlert?s matrix was defined with five levels for feed rate and three levels for cutting speed, according to coded variables. From these experiments the correlation between input and output variables was determined. The best model is in agreement with the analysis of variance and Fischer?s test. Graphs were obtained for both response and contour surfaces. In addition, the optimal point for the best model was determined by the Lagrange criterion, which corresponds to a feed rate (f) of 0.0880 mm rev-1 and the cutting speed (Vc) of 155 mm min.-1, resulting in a minimum roughness average value (Ra) equals to 0.8916 µm.