Enhancing the efficiency of laser beam welding: multi-objective parametric optimization of dissimilar materials using finite element analysis

Laser beam welding (LBW) is a widely acknowledged technology renowned for its efficiency in joining materials through a concentrated laser beam, offering advantages such as rapid processing, high precision, adaptability to diverse materials, low heat affected zone, and seamless integration into automated systems. Despite these merits, achieving precise welds in LBW poses challenges owing to the complicated selection of welding parameters. Additionally, physical experiments encounter difficulties due to high costs, safety risks, and challenges in maintaining control, particularly when welding materials with high thermal conductivity and low melting temperatures, such as AZ31B and AA6061. This research aims to explore the influence of process parameters including laser power, spot size, laser velocity, and segment number, on the determination of residual stress and maximum temperature in LBW. The results of the genetic algorithm (GA) determined optimal input parameters as an average laser beam power of 4011 W, velocity of 30 mm/s, spot size of 0.8 mm, and segment number of 12. The predicted optimal temperature ranged between 1148.2 and 1158.0 °C, while the optimal residual stress fell between 1393.3 and 1534.4 MPa. Additionally, artificial neural networks (ANN) demonstrated results closely aligned with simulated outcomes, affirming the efficacy of the proposed hybrid modeling approaches in optimizing LBW parameters for enhanced welding performance.