Laser beam welding (LBW) uses a concentrated laser beam to fuse materials, providing benefits such as a smaller heat-affected zone (HAZ), speed, accuracy, adaptability across multiple materials, and seamless adoption into automation processes. Achieving optimal weld quality with LBW remains difficult due to the need to select the appropriate welding process variables and proper interlayer between dissimilar materials. Due to the high cost and safety risks to operators and equipment in LBW, it is challenging to use the practical experiment. Additionally, welding materials with high thermal conductivity and lower melting temperatures, such as AZ61A and AA7075, is a considerable challenge. Previous research focused on process parameters in laser beam welding of various metals, but not enough consideration has been given to the effect of a titanium interlayer on AA7075 and AZ61A using virtual data. This study aims to fill these gaps by using simulation data to assess the effects of process parameters and the titanium interlayer during LBW of dissimilar materials. A backpropagation neural network with the gradient descent learning rule was used for optimization, and a central composite design was used to predict the optimal process parameter interaction. The result indicates the optimum equivalent strain is obtained at the values of beam radius between 4.5 to 5 mm. The maximum equivalent stress reached during welding speed is between 4 to 4.5 mm/s. The maximum residual stress was obtained at the number of segments of 160. The predicted maximum and average anticipated errors for peak temperature are 0.1655 % and 0.0210 %, while for residual stress it is computed as 0.1766 % and 0.0754 %. The Ti-interlayer in magnesium and aluminum laser welding reduces peak temperatures, allows for uniform energy distribution, minimizes localized heating, and enhances weld quality while lowering residual stress.
Read full abstract