Response surface methodology for optimization of sintering process for preparation of ZnO low voltage varistor

Optimization of variables in order to maximize an electrical non-linear coefficient (α) of varistor ceramics by traditional method is a daunting and time-consuming task. The objectives of the present work are optimization of sintering temperature and holding time as input variables, and secondly the cooling and heating rates of the sintering process in order to maximize an electrical non-linear coefficient (α) of varistor ceramics. It was achieved by designing the varistor ceramic fabrication using the central composite design (CCD) of four levels of variables, sintering temperature, holding time, cooling rate and heating rate, and a response. Then the varistor ceramics was fabricated in laboratory according to the design to achive actual responses, α. The responses were used for modeling in optimizing the above factors by response surface methodology (RSM). The actual responses were fitted into a valid second order algebraic polynomial equation as quadratic model. The quadratic model was suggested to investigate by analysis of variance (ANOVA) based on some statistical conformation such as lowest standard deviation, P-value, PRESS and the highest Radj, Rpred values. The validated model optimized the above factors by either canonical equation or by 3D plots as optimum point on the surface area. Results indicated that the maximum electric property of the varistor was at temperature 1260°C and holding time of 60 minutes that maximized the nonlinear value around 13. Also, the optimum area is around 5.5 for both cooling rate and heating rate which the maximum α was around 12. Moreover, the model suggested a high desirable solution in special condition to predict the optimum amount of the factors. The condition was including minimum standard error and maximum nonlinearity. The predicted solution included standard error, 0.096, sintering temperature, 1253°C, holding time, 53 min, and α, 11.2. The optimized sample tested by further experiments confirmed that the obtained α value (11) was very close the model predicted value. Then this information was used in optimizing the cooling and heating rates and the predicted solution has standard error of 0.038, cooling rate of 4.65, heating rates of 4.8, and α value of 10.96. The optimized sample tested by experiments, confirmed that the obtained α value is 11.4 was very close the model predicted value. Therefore, RSM was succeeded in modeling of the sintering profile in fabrication of zinc oxide based low voltage varistor.

Preview PDF FS 2014 4 IR.pdf Download (1MB) | Preview

Actions (login required)

View Item