Title: Knowledge base for controlled cost and quality of finite element modelling

Authors: Emmanuel Bellenger, Yassine Benhafid, Nadege Troussier

Addresses: Universite de Picardie Jules Verne, Laboratoire LTI, IUT de l'Aisne, 48 rue d'Ostende, 02100 Saint-Quentin, France. ' Universite de Technologie de Compiegne, Laboratoire ODIC, Centre de recherches Pierre Guillaumat, BP 60319, 60203 Compiegne Cedex, France. ' Universite de Technologie de Compiegne, Laboratoire ODIC, Centre de recherches Pierre Guillaumat, BP 60319, 60203 Compiegne Cedex, France

Abstract: Finite element analysis (FEA) requires the management of simplification assumptions in order to simulate a real complex system. Usefulness of the FEA result highly depends on the pre-processing stage. Thus, defining an appropriate physical model (mathematical model, geometrical simplifications, materials properties, boundary conditions and loads) and an adapted mesh in accordance with simulation goals is a very difficult and time consuming task. In practice, the modelling assumptions are still mostly based on the user|s experience. The paper presents the use and the building of a consultative knowledge base (KB) for FEA. The KB was built mainly by the use of the Design of Experiments method (DOE). The objective of this approach is also to reuse past design and analysis work as analysis template. From the reuse of previously validated analysis components, the automated method supports the selection of the physical model and adapted mesh to be simulated in order to control the cost (time) and the quality of the whole FEA.

Keywords: finite element analysis; FEA; knowledge base; reusability; design of experiments; DOE; modelling; mesh generation; cost control; quality control; design reuse; simulation.

DOI: 10.1504/IJDE.2008.021169

International Journal of Design Engineering, 2008 Vol.1 No.2, pp.166 - 191

Published online: 07 Nov 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article