Abstract—Robu1st design experiments with single array were applied for the first time to optimization of rubber examination glove production in Malaysia. Seven controllable factors and one noise factor were combined in a single experimental setup at fixed levels also called a combined array or a single array. Two levels saturated fractional factorial designed was employed to elucidate the controllable factors that significantly affect the examination glove. The inclusion of the noise factor (relative humidity) allows a systematic study of the effect of process variation on product and process quality that can be expected under processing condition. The modeling was performed on the mean and standard deviation separately for the quality characteristics of interest (pinholes). The standard deviation model is used to investigate the possibility of reducing the process variations induced by uncontrolled factors while keeping the mean on target. In this paper, we demonstrate how response modeling derived from robust design experiments were employed to identify setting of the controllable factors so that quality characteristics are least sensitive to variation of noise factor.

Index Terms—Design of experiments, robust design, noise factor, multiple regression.

Z. Wahid is with the Department of Science in Engineering, Faculty of Engineering, International Islamic University Malaysia (e-mail: zaharahwahid@iium.edu.my).M. T. Tham is with Department of Chemical and Process Engineering, Faculty of Engineering, University of Newcastle.

Cite: Zaharah Wahid and Ming T. Tham, "Single Array and Response Modeling of Robust Design Experiments," International Journal of Applied Physics and Mathematics  vol. 2, no. 5, pp. 359-361, 2012.