Title: Improving foundry process control: an investigation of cluster analysis and path model

Authors: Vivek V. Khanzode, J. Maiti

Addresses: KITS College of Engineering, Maharastra 416122, India. ' Department of Industrial Engineering and Management, IIT, Kharagpur 721 302, India

Abstract: Although the use of multivariate statistical methods in quality control is numerous, the relevance and use of cluster analysis and path modelling in multivariate process control have not been effectively investigated. In this study, an attempt was made to model and control the melting process of a grey iron foundry in western India using cluster analysis and path modelling. Online data was collected for 18 variables in total: three raw materials, six processes, and nine quality variables. Initially, cluster analysis was done and the raw material, process and quality variables were grouped into four clusters. Then, path modelling was conducted to model the interrelationships of the quality, process and raw material characteristics of each of the four clusters separately. The results of these analyses were used to develop an improved control scheme and validated through field investigation under the control scheme. The study showed encouraging results with substantial improvement in Cp values for the quality parameters.

Keywords: grey iron foundry; heat duration; melting process; multivariate control schemes; process parameters; process control; cluster analysis; path modelling; India; quality control; quality management.

DOI: 10.1504/IJPQM.2007.013335

International Journal of Productivity and Quality Management, 2007 Vol.2 No.4, pp.404 - 422

Published online: 21 Apr 2007 *

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