| || Design and Application of Economical Process Control Charts
Author : Rambabu, Y. ;Murthy, S.S.N.
Source : Defence Science Journal ; Vol:47(1) ; 1997 ; pp 45-53
Subject : 519.2 Statistics
Keywords : Optimum control parameters ;Fibonacci search;Statistical quality control
Abstract : Statistical quality control techniques are useful in monitoring the process behaviour. Attribute control charts are widely used in process control. The selection of sample size, sampling interval and control width of the control chart is important in minimising the quality costs. Control chart parameters, like 3delta controlimits and fixed fraction sampling at conveniently selected sampling intervals result in deplovable cost penalties in quality control. The best selection of these parameters depends on several process parameters,like frequency of occupancy of a shift in the process, cost of sampling, cost of investigation for finding assignable cause, probability of false alarms, penalty cost of defectives and process correction cost. A general model has been developed to determine the total quality cost as a function of these parameters. Probability of not identifying a process shift (Beeta-risk) and probability of wrongly. concluding the process got shifted (alpha-risk) are considered in developing the model. This cost equation is optimised to determine optimum values control chart parameters. Fibonacci search is used to quicken the analytical method or determining optimum sampling size and control width. The proposals made by Duncan, Montgomery, Gibra and Chiu for determining the optimum control chart parameters are critically examined and compared with the present model. Case studies were conducted in two foundries. Optimum control chart parameters in casting of cylinder liners and cast plates are determined. It has been found that quality costs are considerably reduced by using optimally designed control chart parameters with proposed method.