Title: A hybrid system for component selection experience with bearing selection

Authors: Kamel Mehdi, Ali Zghal

Addresses: Institut Preparatoire aux Etudes d'Ingenieur El Manar, BP 244, 2092 Tunis, Tunisia. ' Ecole Superieure des Sciences et Techniques de Tunis, 5 avenue Taha Hassen BP 56 Bab Mnara 1008 Tunis, Tunisia

Abstract: More and more modern mechanical systems are assembled using prefabricated components. Optimal selection of these components has a strong effect on the quality and price of the system. The prerequisites of the selection are having access to the different component data selection, knowledge and cooperation of various softwares. This paper presents a hybrid system applied to mechanical engineering design for bearing selection. The system is based on the definition of the different bearing classes with object oriented language and a knowledge database for multiple decisions for a given application. The system is an interactive software package, ROULEXPERT that assists the designer in selection the correct combination of ball and roller bearings to support shaft for a given set of operating conditions. The system executes dynamically different methods and procedures to evaluate and extract characteristic data of the selected bearing from database. The selection procedure is demonstrated by an example.

Keywords: computer aided design; CAD; artificial intelligence; geometric modelling; hybrid systems; expert systems; object-oriented language; knowledge based systems; KBS; mechanical engineering design; component selection; bearing selection; ball bearings; roller bearings.

DOI: 10.1504/IJVD.2005.007221

International Journal of Vehicle Design, 2005 Vol.39 No.1/2, pp.80 - 92

Published online: 24 Jun 2005 *

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