Gráfico de controle modificado para monitoramento de variância

Abstract

Control charts are powerful tools used by many industries to monitor the quality of their processes and detect special causes of variation. They are often used to monitor the mean of some process quality characteristic with the well known X ̅, while the process variability can be monitored with either a control chart for the range, called R control chart, for the standard deviation, called S control chart or for the variance, called S^2 control chart. This work will focus on the S^2 control chart. In their original formulation, if the actual process mean (μ) or variance (σ^2) are different or larger from their specified in-control values (μ_0 and σ_0^2, respectively), the process is declared out-of-control. However, in many practical situations, even though the process may be declared out-of-control, it might be still capable from a practical point of view in terms of the proportion of nonconforming items produced. Thus, it may not be necessary to stop the process and start looking for assignable causes, which can save time and resources. The Modified Control Charts were designed to monitor the process mean in such a capable situation. However, regarding the monitoring the process variance, there is still no control chart presented in the Statistical Process Control (SPC) literature that considers the capability of the process while monitoring the process variance. According to Juran and Godfrey (1998), capability is the sense of a competence, based on tested performance, to produce quality products. With this background as motivation, in this work, we derive a Modified Control Chart developed to monitor the mean, for monitoring the process variance, considering the capability of the process.