Modeling forced convection in the thermal simulation of laser cladding processes

Abstract

A comprehensive methodology for the implementation of thermal convection into the finite element (FE) analysis of laser direct energy deposition (DED) cladding is developed and validated. Improved convection modeling will produce improved thermal simulations, which will in turn yield more accurate results from subsequent models seeking to predict microstructural changes, deformation, or residual stresses. Two common convection implementations, considering no convection or free convection only, are compared to three novel forced convection methods: forced convection from heat transfer literature, forced convection from lumped capacitance experiments, and forced convection from hot-film anemometry measurements. During the cladding process, the exposed surface, the surface roughness, and total surface area change due to material deposition. The necessity of accounting for the evolution of the mesh surface in the FE convection model is investigated. Quantified error analysis shows that using any of the three forced convection methodologies improves the accuracy of the numerical simulations. Using surface-dependent hot-film anemometry measured convection yields the most accurate temperature history, with L 2 norm errors of 6.25−22.1 ∘C and time-averaged percent errors of 2.80–12.4 %. Using a physically representative convection model applied to a continually evolving mesh surface is shown to be necessary for accuracy in the FE simulation of laser cladding processes.