Title: An alternative to Monte Carlo simulation method

Authors: Mohamed H. Gadallah

Addresses: Department of Operations Research, Institute of Statistical Studies and Research (ISSR), Cairo University, Giza, Cairo 12613, Egypt

Abstract: Tolerance analysis usually complements tolerance synthesis; however, analysis algorithms, especially the Monte Carlo method proves highly expensive. This study offers an alternative to the Monte Carlo method. The huge sample sizes are replaced by small orthogonal experiments. The mean and variance of critical design requirements are calculated via the two methods. Preliminary results indicate the need for larger size orthogonal arrays to conform to results by Monte Carlo simulation (MCS). Two new modifications are introduced and their effects on the optimum tolerances and process combinations are studied. Results indicate the potential of the algorithm for further future developments. This study claims two important things: a) not all variables are of the same importance in their domains b) there is a reasonable alternative to Monte Carlo simulation method subject to use of larger size arrays.

Keywords: Monte Carlo simulation; approximations; design of experiments; DOE; mechanical design; orthogonal arrays; tolerance analysis; variables.

DOI: 10.1504/IJEDPO.2011.040261

International Journal of Experimental Design and Process Optimisation, 2011 Vol.2 No.2, pp.93 - 101

Published online: 11 Oct 2014 *

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