AbstractThe low process stability of laser welding of copper with near-infrared lasers requires precise input data for process control and meaningful simulations. But meanwhile, available datasets of temperature-dependent reflectance or absorptance for near-infrared lasers on copper do not show good agreement between the different sources and often do not include the fusion process, which is of crucial importance for realistic laser welding simulations. Additionally, most of the datasets are only calculated. Therefore, in a previous study, temperature-dependent reflectance measurements were performed on electro tough-pitch copper using a near-infrared laser. The measurements revealed a reflectance drift, which was induced by the setup behavior during heating, and the time-dependence of chemical reactions like the redox-reaction as possible error sources. In this study, experiments on laser melting as the fundamental process of laser welding were performed, together with corresponding simulations using the measured reflectance values for oxide-reduced and for untreated copper from the previous study. Then, the simulations were compared with the experiments to estimate the accuracy of the reflectance measurements. To provide context, the same simulations were also conducted using reflectance datasets from other authors. In a second step, the reflectance data were corrected with respect to the reflectance drift and the effects of redox reactions were adapted to the conditions of the laser melting experiments. Using the resulting reflectance curves, an improved agreement of simulation results and the experiments was achieved over a range of different test cases, without the necessity of correction factors in the simulation model.
Read full abstract