A neural network approach for selection of Powder Metallurgy process parameter of rapidly solidified white cast iron

Powder Metallurgy (P/M) involves multiple input and output which are non-linearly related for which statistical optimization methods are not suitable. These considerations lead to adoption of neural network (NN) for proper selection of P/M process parameter. In the present work, white cast iron powder is taken as the work material and NN approach is employed which allows specification of multiple input and generation of multiple output recommendations. The NN model developed for the purpose is based on three-layer resilient back propagation learning algorithm with the help of Matlab NN Toolbox. Supervised training has been adopted to train the network which helps in prediction of process parameter such as sintered density and growth % at different compaction pressure and sintering temperature of consolidated water atomized rapidly solidified white cast iron. Training data are collected by the experimental setup in laboratory. The density and growth % predicted by NN model coincides well with the experimental data with a tolerable error of 1 × 10^-10 which confirms its capability over the standard design procedures. The mathematical model developed here can be used as knowledge based system for a number of P/M products.