Prediction of Shot Peen Forming Parameters of Integral Aircraft Wing Panels

Xianjie Zhang,Junbiao Wang,Yongjun Wang,Mingjie Qiao
DOI: https://doi.org/10.4028/www.scientific.net/msf.532-533.937
2006-01-01
Materials Science Forum
Abstract:In this paper, algorithms are presented for predicting peen forming parameters of integral aircraft wing panels with complex airfoil shapes. The peen forming deformation is divided into stretching deformation and bending deformation. The stretching deformation is assumed to result from the tensile strain within the plane panel, and the bending deformation corresponds to the difference of maximum and minimum curvature of the airfoil surface. The distribution of the forming tensile strain within the panel is obtained by optimal mapping of the airfoil surface in the sense that the stretching deformation energy from the plane panel to its spatial shape is minimized. In order to fit the nonlinear relation between the peening parameters and the deforming parameters, a back-propagation (BP) artificial neural network (ANN) is modeled with input parameters of thickness, curvature, tensile strain, etc, to predict the peening parameters of coverage, air pressure and intensity. Experimental peen forming data are given to train the BP ANN. It’s verified that the predicting methods are effective.
What problem does this paper attempt to address?