Two step multivariate modeling and optimization of sintering profile of electrical ceramic fabrication process to enhance the electrical protection of the electronic appliances

摘要

Globally, billions of electronic devices are discarded due to their malfunction of low electrical protection from a non-linear property ceramic core varistor. The common varistor ceramic microstructure is made by the pressed powder of ZnO and small amount of additives such as Bi 2 O 3 , TiO 2 , Co 3 O 4 , Mn 2 O 3 , Sb 2 O 3 and Al 2 O 3 . The non-linearity is originated from the microstructure that is fabricated by a sintering profile. The profile is included sintering temperature, holding time, heating and cooling rate. In this work, the profile was modeled and optimized by multivariate method to enhance the non-linearity consequently improve the electrical protection. Therefore, two series of experiments were designed and were performed in laboratory to obtain the actual non -linear coefficient (α). The designs consisted of the sintering components and α as input variables and output response respectively. The actual results were used for two steps modeling of the ceramic fab rication's sintering profile. In the first step, the temperature and holding time of the sintering process were considered as input effective variables while heating and cooling rates were kept constant at 5 .C/min. However, the input of the process in the second step were heating and cooling rates at optimized temperature (1253 o C) and holding time (56 min). The outputs of the both steps were the calculated α that obtained from electrical characteristic of the fabricated ceramic. The results of performed design were used to model of the fabrication which was validated by analysis of variance. The validated model was used for determination of optimum value o f the input variables that maximized α. Moreover, the model predicted a desirable condition of ceramic fabrication that was experimentally validated and used as final varistor ceramic. The ceramic was characterized by I-V characteristic to calculate α. The calculated α was 35.22 in first step while it was improved to 42.18 in the second step of modeling and optimization. In conclusion, the multivariate modeling and optimization which could be industrial scale up has been succeeded to promote the protection of electric and electronic devices.