A hybrid-constrained MOGA and local search method to optimize the load path for tube hydroforming

摘要: The production of a tubular hydroformed part often requires combination internal pressure and axial force at the tube ends to fully form its specified geometry. A successful hydroforming process not only achieving that conforms design specifications, but also ensures has reasonably uniform thickness distribution is free defects, such as wrinkles, severe thinning, or fractures. load path (pressure vs. end feed history) largely determines robustness quality finished parts. In this paper, hybrid constrained optimization method was proposed solve type multi-objective problem by coupling genetic algorithm local search. procedure developed considering five objectives: four formability objectives (i.e., minimize risk wrinkling, global fracture) geometric objective (to corner radius). Kriging predictor used accelerate computation operations generate new feasible solutions. Finite element simulations were after each generation accurately evaluate offspring, solutions with rank 1 retained throughout all generations. Once Pareto obtained algorithm, search carried out in regions interest assistance visualization. This applied straight create an expanded region square cross section; optimum produced very safe radius 9.115 mm maximum thinning 23.9%.