Phosphate, phosphide, nitride and carbide capacity predictions of molten melts by using an artificial neural network approach

In the present study, the impurity capacities (Ci) of phosphate, phosphide, nitride and carbide in binary and multi-component molten melt systems at different temperatures were estimated using the artificial neural network approach. The experimental data taken from the previous studies were introduced to the artificial neural network, then the calculated results were plotted against the experimental values for comparative purposes. Besides, iso-phosphate capacity contours on the liquid region of CaO-CaF₂-Al₂O₃ ternary phase diagram at 1 773 K were generated and plotted by using the neural network model results. The calculated results obtained through neural network computation agreed well with the experimental ones and were found more accurate than those estimates based on some models.