Title: Optimising the Gaussian-distributed filling process parameters

Authors: Mohammed H. Hassan

Addresses: Department of Mechanical Engineering, IE Division, Helwan University, Helwan, Cairo, Egypt

Abstract: A filling production system is that characterised by two contradicting nonconforming cost components. The first component is that incurred when the product specification exceeds the upper tolerance limit, called an overfill cost, whereas the second cost component is contributed when the product specification falls below the lower tolerance limit, called an underfill cost. The ratio of the two cost components affects to a high extent the choice of the process parameters, especially the process location parameter that minimises the overall costs. The work in this research is devoted to finding the filling process|s optimum parameters; namely the process|s mean and variance, considering a Gaussian-distributed process. Firstly, the process|s variance is considered constant; and searching for the optimal process mean parameter takes place. An analytical, closed form, solution model is developed that realises such target. The developed model is ended up with a graph that can simply be used by concerned engineers and practitioners. The assumption of a constant process variance is then relaxed; and the developed model is extended to involve the variance reduction cost. An algorithm is introduced to solve the extended model that optimises the filling process| mean and variance.

Keywords: filling process; optimisation; overfill cost; underfill cost; variance reduction cost; error function; manufacturing.

DOI: 10.1504/IJASM.2008.019605

International Journal of Agile Systems and Management, 2008 Vol.3 No.1/2, pp.163 - 174

Published online: 17 Jul 2008 *

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