As a critical component of the propellant tank, the tank bottom is subjected to complex loads such as internal pressure and vibration and has high requirements for structural load-bearing capacity. Hydroforming deep drawing is one of the techniques for the integral forming of the tank bottom. As the tank bottom is a large-size thin-walled structure, defects such as cracks and wrinkles are prone to occur during the hydroforming deep drawing process. Aiming at reducing these defects, the hydraulic pressure loading path and blank holder force loading path of the hydroforming deep drawing process are studied, and a multi-objective optimization method is proposed to improve the surface accuracy and thickness distribution uniformity of the tank bottom. The complex loading path curve optimization problem is transformed into a functional relationship between hydraulic pressure and blank holder force with time. The hydraulic pressure and blank holder force at each time node are used as design variables, and the maximum wall thickness reduction rate, rupture trend factor, wrinkle height, and wrinkle trend factor are used as optimization targets. The radial basis function (RBF) neural network is used to establish the approximate model between the loading path and the optimization target, and the multi-objective particle swarm optimization (MOPSO) algorithm is used to optimize the solution. Taking the hemispherical tank bottom as an example, the optimal hydraulic pressure loading path and blank holder force loading path are obtained, and the quality of the formed part is improved.
Read full abstract