Mathematical Modelling of Vickers Hardness of Sn-9Zn-Cu Solder Alloys Using an Artificial Neural Network

An artificial neural network (ANN) model was used for the simulation and prediction of the mechanical properties of Sn-9Zn-Cu solder alloys. Sn-9Zn-Cu solder alloys containing different Cu contents (0, 1, 2, 3, 4 and 5 wt%) were successfully prepared by permanent mold casting. The specimens were heated in a protective argon atmosphere at 433 K for 24 h, followed by water quenching at 298 K. Finally, the heat-treated samples were aged at 373 K for different time intervals (t(a) = 2, 4, 8, 16 and 32 h), followed by water quenching at 298 K. The phases present in the current alloys were detected by X-ray diffraction analysis. For morphological characterization, a scanning electron microscope operated at 20 kV was tilized. The mechanical properties of the samples were studied using hardness measurements. The variations in the hardness data with increasing aging time were determined based on the structural transformations that take place in the alloys. The ANN model was applied to the hardness measurements to simulate and predict the Vickers hardness of Sn-Zn-Cu alloys with mean square error values equal 9.55E-06 and 9.44E-06 for training and validation data respectively after 281 epochs. The simulated and predicted results were consistent with the experimental results.

Automated summary

An artificial neural network (ANN) model was utilized to simulate and predict the mechanical properties of Sn-9Zn-Cu solder alloys, showing consistent results with experimental data, indicating the good accuracy and reliability of the ANN model in predicting alloy hardness.