WEAR RATE PREDICTION OF FRICTION STIR WELDED DISSIMILAR ALUMINUM ALLOY BY ANN

The effect of process parameters on the mechanical properties of dissimilar AA6061 t6–AA5083 0 joints produced by friction stir welding was studied. Different samples were produced by varying the advancing speeds of the tool as 20 and 40 mm/min and by varying the alloy positioned on the advancing side of the tool. The rotating speed is varied from 600 to 900 rpm. Taking all these above-mentioned points into consideration, it is obvious that the relationship between these parameters has to be established in order to understand their interdependencies. Artificial neural networks possess such predictive capabilities, when trained, to understand the relationship between the dependent and independent variables of the system. This work involves a neural network with 4 independent variables. The aim is to train the network so that it will be able to predict with reasonable accuracy. In order to use the network for prediction, it must show the least possible root mean square error during validation