Joint Stiffness Identification of Body Structure Using Neural Network. For T-shape Jointed Box Beams.

Abstract

Automobile body structures are constructed from spot-welded thin-walled box beams. Joint stiffness of body structure is influenced by the plate thickness, service hole, spot welding around the joint, and the shape of the cross section of box beams. In this paper, an application of hierarchical neural networks to joint stiffness identification of body structure is described. First, sample data of effective parameters vs joint stiffness are calculated by the finite-element method. Second, the error-back-propagation neural network is trained using the sample data. It is found that the value of joint stiffness for effective parameters can be obtained by the trained neural networks.