Accuracy of calculated component distortions using the weld pool length to calibrate the heat source

A better understanding of the laser welding process is fundamental for improving the quality of welded components. Hence, an experimental analysis, as well as an increase of the accuracy of computational approaches is important. When using a structural simulation to predict the component distortions caused by a welding process, it is necessary to reconstruct the temperature field. The use of heat source models is an established approach for the reconstruction of the temperature. This work is focused on evaluating the accuracy of calculated component distortions using the weld pool length to calibrate the heat source. For the implementation, image processing tools and an automated algorithm to calibrate the heat source are used. The experiments were carried out using an adapted intensity distribution which results from superimposing the beams of a high-power diode laser and a neodymium-doped yttrium aluminum garnet (Nd:YAG) laser. To observe the process in general and the weld pool in particular, a high-speed camera was used. The image data was subsequently analyzed by an algorithm to extract the transient weld pool length. To simulate the laser welding process in an FEA-tool, a parameter calibration of two double ellipsoid heat sources, based on optimization algorithms, was used. To compare the computational and the experimental results, the component distortions were measured by position sensors. The results of the study are shown and the relevance of implementing the weld pool length to calibrate the heat source is pointed out.