Prediction of Bending Angle for the Laser Forming of Titanium Sheet Metal
K. Paramasivan, S. Das, M. Sundar and D. Misra
Laser bending of metals and alloys have good potential in automotive and aerospace industries; however, the nonlinear and unpredictable nature of the bending process limits its applicability. This paper aims to develop a hybrid model based on finite element (FE) analysis and artificial neural network (ANN) to predict and control the bending angle of titanium sheet during laser bending. A three-dimensional (3-D) numerical model with moving heat source is developed using FE formulation. The results of the FE analysis are linked with a back propagation neural network model. Results show that the ANN model can predict the bending angle with a good accuracy compared to other models.
Keywords: Laser bending, finite element (FE), artificial neural network (ANN), titanium, bend angle, numerical model, response surface methodology (RSM)