| Summary: | Artificial neural network (ANN) and Response Surface Methodology based on a 25â1 fractional factorial design were used as tools for simulation and optimisation of the dissolution process for Azaraegbelu clay. A feedforward neural network model with LevenbergâMarquard back propagating training algorithm was adapted to predict the response (alumina yield). The studied input variables were temperature, stirring speed, clay to acid dosage, leaching time and leachant concentration. The raw clay was characterized for structure elucidation via FTIR, SEM and X-ray diffraction spectroscopic techniques and the result indicates that the clay is predominantly kaolinite. Leachant concentration and dosage ratio were found to be the most significant process parameter with p-value of 0.0001. The performance of the ANN and RSM model showed adequate prediction of the response, with AAD of 11.6% and 3.6%, and R2 of 0.9733 and 0.9568, respectively. A non-dominated optimal response of 81.45% yield of alumina at 4.6Â M sulphuric acid concentration, 214Â min leaching time, 0.085Â g/ml dosage and 214Â rpm stirring speed was established as a viable route for reduced material and operating cost via RSM. Keywords: Alumina dissolution, ANN modelling, Azaraegbelu, Clay, RSM
|
|---|