Multi-objective topology optimization of deep drawing dissimilar tailor laser welded blanks; experimental and finite element investigation

Laser welded blanks (LWBs) are semi-finished components typically manufacture by dissimilar materials, thicknesses, shapes, coatings, etc. After butt welding of the primary sheets, the product sheets are subjected to the sheet metal forming process. Formation of the heat-affected zones (HAZ) is typical in LWBs, which possess quite different mechanical properties than the base materials. Recently, laser beam technologies have been widely employed to weld different types of vehicles panels. In this study applying Nd:YAG laser welding, experimental and numerical investigations are carried out to evaluate the effects of process input factors on deep drawing process of LWBs. Laser beam power, welding speed, blank holder force (BHF), material properties, and friction coefficient are considered as process key input parameters. In addition, the laser welding and deep drawing processes were numerically simulated using Simufact Welding and Abaqus/Explicit software, Used the Simorgh supercomputer for heavy modeling calculations. Moreover, drawing depth, weld line movement, and energy absorption are taken into account as process main outputs or objective functions. Besides, using an advanced MATLAB code, multi objective optimization based on genetic algorithm is applied to determine the optimal design input parameters. It is observed that the critical stresses were taken place outside the weld zone and rupture due to high heat input of laser and metallurgical changes of the base metal occur in the pre-softening zone. In addition, the weld line displacement occurs as a result of plastic strain change of the weld joint that causes failure-prone zone creation as well as the adverse wrinkling. By considering weld line displacement and absorbed energy as multi-objective function, the optimal points is 1.15 mm and 0.21 KJ for weld line displacement and absorbed energy, respectively. Good agreement between the simulated and the experimental results revealed that the model would be appropriate for deep drawing of LWB process numerical simulation.