Please use this identifier to cite or link to this item: http://localhost:80/xmlui/handle/123456789/146
Title: A Superstructure of Process Capability Indices for Circular Specification Region
Authors: Chatterjee, Moutushi
Chakraborty, Ashis Kumar
Keywords: Circular specification region
Elliptical process region
Bi-variate process capability indices
Superstructure
Threshold value
Proportion of non-conformance
Issue Date: 5-Jan-2012
Series/Report no.: Communications in Statistics—Theory and Methods;44: 1158–1181, 2015
Abstract: Circular specification regions can be seen in processes like hitting a target (in ballistics), drilling a hole (in manufacturing industries) and so on. However, only a few process capability indices are available in the literature to address the problem. Most of these indices, in turn, make some assumptions like equality of variance and independence of the two axes of the circular tolerance region. Since, in most of the cases, these assumptions are not practically viable, in the present article, we have proposed a few of the process capability indices which do not need the above assumptions to be valid. Also, we propose a superstructure which unifies all the proposed indices. Some properties of these indices have been studied including the threshold value and the relationship of the proportion of non-conformance with the member indices of the superstructure. These strengthen the practical utility of the superstructure. Distributional properties like expectations and variances of the member indices of the superstructure are also studied to have a better insight about the indices. A real life example has been discussed to carry out a comparative study of the performance of the existing as well as the newly developed indices.
Description: Communications in Statistics - Theory and Methods
URI: http://hdl.handle.net/123456789/146
ISSN: 0361-0926 print
1532-415X online
Appears in Collections:Journal Article

Files in This Item:
File Description SizeFormat 
Moutusi - circular (1).pdf673.12 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.