Surrogate model of the thermoforming of fiber reinforced thermoplastics

Abstract. In the thermoforming process shearing of fiber reinforced thermoplastics (FRTP) leads to wrinkle formation and is therefore one most critical deformation mechanism. The target value predict shear angle which considered in this paper. To distribution during FRTPs a Convolutional Neural Network (CNN) used. approach based on consideration principal curvature characteristics, resulting reduction any complex geometry planar, parabolic, elliptical, semi-spherical hyperbolic areas. similar research, CNNs are trained for each specific geometry. work, CNN refers database mentioned characteristics. theory, all forming geometries can be created from these, why enabled result paper, benchmark double dome selected as validation error prediction compared with results numerical simulations performed.