A New Local Contact Search Method Using a Multi-Layer Neural Network
Atsuya Oishi; and Shinobu Yoshimura

doi:10.3970/cmes.2007.021.093
Source CMES: Computer Modeling in Engineering & Sciences, Vol. 21, No. 2, pp. 93-104, 2007
Download Full length paper in PDF format. Size = 242,010 bytes
Keywords Contact Search, Neural Network, Finite Element Method, Sliding Interface, Curved Surface.
Abstract This paper describes a new local contact search method using a multi-layer neural network and its application to smoothed contact surface consisting of Gregory patches. A contact search process consists of two phases: a global search phase for finding the nearest node-segment pair and a local search phase for finding an exact local coordinate of the contact point within the segment. In the present method, the multi-layer neural network is utilized in the latter phase. The fundamental formulation of the proposed local contact search method is described in detail, and it is applied to smoothed contact surfaces consisting of Gregory patches. Through sample analyses, it is proved that the proposed method is fast and accurate enough for practical applications. Specifically, it is several times faster than the conventional method in the case of smoothed contact surfaces.
PDF download PDF