Abstract In the literature on numerical simulations of Marangoni driven hydrodynamics in laser spot weld pools, it has been proven impossible to obtain good agreement between simulated and experimentally observed weld shapes without artificially enhancing the thermal conductivity and the dynamic viscosity of the liquid steel by one to two orders of magnitude. This has mostly been ascribed to flow instabilities that are not properly accounted for in the simulations. However, whereas experimental weld shapes are obtained post solidification, the cooling and solidification stage is generally neglected in reported simulations. In the present work, we report a detailed study on the role of the artificial diffusivity enhancement factors in weld pool simulations, and we extend the simulations into the cooling and solidification stage. We show that during the cooling stage, flow reversal may occur in the weld pool, which enhances the downward heat and momentum transfer. This leads to a deeper weld pool that agrees better with experimental results. By including the cooling and solidification stage into the simulations of weld pool hydrodynamics, an improved agreement with experimentally observed weld shapes can be obtained with a reduced necessity to artificially enhance the thermal conductivity and the dynamic viscosity@en

Various physical interfacial phenomena occur during the process of welding and influence the final properties of welded structures. As the features of such interfaces depend on physics that resolve at different spatial scales, a multiscale and multiphysics numerical modeling approach is necessary. In a collaborative research project Modeling of Interface Evolution in Advanced Welding, a novel strategy of model linking is employed in a multiscale, multiphysics computational framework for fusion welding. We only directly link numerical models that are on neighboring spatial scales instead of trying to link all submodels directly together through all available spatial scales. This strategy ensures that the numerical models assist one another via smooth data transfer, avoiding the huge difficulty raised by forcing models to attempt communication over many spatial scales. Experimental activities contribute to the modeling work by providing valuable input parameters and validation data. Representative examples of the results of modeling, linking and characterization are presented.@en