Application of Intelligent Techniques for Controlling the Green Sand Properties Application of Intelligent Techniques for Controlling the Green Sand Properties

In the present work an Artificial Neural Network model and a Neuro-Fuzzy model were developed for predicting the properties of clay bonded moulding sand mix. Experiments have been conducted to generate the data for modelling. Various sand mixes have been prepared varying the amounts of clay, coal dust, moisture and mulling time. For each sand mix composition, the permeability was measured. The models were trained and tested with data set and were used for predicting the properties of moulding sand mixture. The predicted values of the moulding properties obtained by the two models were found to be in good agreement with the experimental values. The results of the neuro-fuzzy model were compared with the results obtained from the neural network model. It has been observed that neuro-fuzzy model predicts the properties more accurately as compared to neural network model. Individual foundries may generate their own green sand property data using their own raw materials and predict the properties for different sand compositions using such intelligent techniques.