Computational modeling of laser welding for aluminum–copper joints using a circular strategy

Abstract

Dissimilar metal welding poses a significant challenge for researchers in the welding and materials science communities due to the different properties of the metals involved. In this study, we present a novel approach for determining the optimal processing parameters for continuous laser welding (CLW) of dissimilar metals, specifically aluminum and copper, using a circular strategy. The approach utilizes computational fluid dynamics (CFD) simulations to model the welding process and predict joint quality. We validated a CFD-based numerical model for predicting the melt-pool shape across a range of laser power and circular welding speed values. Additionally, we developed artificial neural network (ANN) models that can predict melt pool interface width, penetration depth, and copper concentration for any combination of processing conditions. Using surrogate modeling, we identified the optimal values of laser power and welding speed that result in the highest quality joint. Our results showed that the optimal processing conditions were a laser power of 2950W and a circular speed of 55 mm/s, which resulted in a high-quality Al–Cu joint with maximum shear strength of 2200 N. Microstructural analysis and tensile-shear strength testing confirmed that the optimal conditions resulted in a low degree of intermixing and the absence of large intermetallic compounds (IMCs) in the weld seam. Overall, study offers a valuable contribution to the welding and materials science communities by presenting a novel approach to address a common challenge in dissimilar metal welding. Our approach can be applied to other types of dissimilar metal welding and may inform the development of new welding techniques in the future.

Keywords

• Continuous laser welding (CLW)
• Computational fluid dynamics (CFD)
• Artificial neural network (ANN)
• Parameter optimization
• Intermetallic compound (IMC)