Springback Characteristics of Cylindrical Bending of Tailor Rolled Blanks

As a new type of variable thickness sheet structure, the TRB (tailor rolled blank) has good prospects for the development of lightweight materials in the automotive industry. However, springback is a key issue in production. Research on TRB springback characteristics has great significance for further applications due to variations in the sheet thickness and gradient distribution of the material mechanical properties. In this study, the springback characteristics of TRBs were investigated by means of the finite element code ABAQUS/USDFLD and experiments taking cylindrical bending as an example. The results showed that the cylindrical bending process of the TRB gradually evolved from three-point bending to four-point bending and, finally, to multipoint bending. At same time, the gradient of the thickness leads to the nonuniform longitudinal distribution of the von Mises stress. On the contrary, larger bending angles can be achieved by reducing R and improving Rd, but t/T has little effect on the bending angles. In terms of the influence of springback, increasing Rd and reducing R and t/T can lead to a smaller springback angle. This project provided an important opportunity to advance the understanding of TRB springback characteristics.

Convolutional Neural Network Using Bayesian Optimization for Laser Welding Tailor Rolled Blanks Penetration Detection

Convolutional neural network (CNN) is an efficient and robust method which can accurately detect the Tailor Rolled Blank laser welding pool penetration status. To select proper hyperparameters and optimization of CNN model are black box problem. In this paper, an innovative method based on CNN to identify the penetration status of the weld pool during laser welding was introduced. A coaxial monitoring platform is set up, as well as two-class, three-class and four-class datasets are created for training and validating the CNN. The Bayesian Optimization (BO) method is used to optimize hyper-parameters which are adopted for training CNN model, determine the best parameters of depth, initial learning rate, momentum and L2 regularization. The results show that using BO method leads to accuracy improvement compared with the CNN model trained from scratch with default hyper-parameters, hence it can effectively solve the problem that the hyper-parameters of CNN are difficult to adjust. Under various laser welding parameters, high-accuracy detection of penetration status can be acquired with the test accuracy of four-class reaching 95.2%, which slightly lower than the test accuracy of the three-class and two-class.