Prediction of Mechanical Properties of Friction Stir Processed AZ31 Using Neural Network

Friction Stir Processing (FSP) of AZ31 with nano-sized TiN particulates was employed to produce AZ31/TiN nano-composite, and hardness values were measured using Vichers indenter. An Artificial Neural Network (ANN) was applied to acquire the relationships between hardness and the processing parameters of FSP. In establishing of those relationships, tool rotation & substrate advancing speeds and number of FSP passes were taken as the inputs, whilst the hardness was presented as the output. The network was trained using the data from experimental work. It was found that the trained neural network could predict the hardness quite reliably and the optimum processing parameters can be quickly selected to achieve the desired hardness using the prediction based on the ANN model.

Keywords:

Friction stir processing , Hardness , Artificial neural network , Nano-composite