modeling and optimization of penetration depth using genetic algorithm, focusing on submerged arc welding

Abstractsubmerged arc welding is a production process directly affected by various input parameters and interactions, and these effects will directly affect the output results. this research investigated the impact of arc voltage, electric current intensity, electrode stick-out, welding speed, and the thickness of the layer of nanoparticles on the penetration depth. in this study, central composite rotatable design and analysis of variance were used to examine the effect of input variables on output results and reduce time and costs. ultimately, the genetic algorithm was used to find the optimal levels of selected parents. the results indicate that increasing the arc voltage and the electric current intensity due to more input heat transfer caused more melting of the base metal and thus increased the penetration depth. the difference in thermal conductivity between the base metal and the nanoparticles results in less heat transfer to the inner layers of the workpiece; therefore, the high thickness of the layer of nanoparticles reduces the penetration depth. on the other hand, increasing the thickness of the layer of nanoparticles increased the weld’s height and width.