| Summary: If the values of the quality variables of the products produced in a production process deviate from the predetermined standards, the prevention of the losses that may occur by interfering with the process is of great importance in terms of both time and cost. In this respect, it is obvious that the studies to be carried out by using statistical process control methods to keep the production processes under control will make significant contributions to the enterprises. In this study, the control status of the production process of Medium Density Fiberboards produced in a board manufacturing enterprise operating in the forest products industry and the success of the process in meeting consumer specifications was investigated in terms of some quality variables. For this purpose, the collected data of the variables were examined by using multivariate Hotelling T2 and Generalized Variance and univariate ̅ and R charts. In addition, the process capability analysis was utilized to detect the degree of meeting customer specifications of the process. The T2 values of the signals formed in the Hotelling T2 chart were decomposed by Mason-Young-Tracy method and it was determined which quality variable contributed more to each signal. As a result; it is detected that the process was not in control for the ̅ and Hotelling T2 charts, which reveal the shift level in the mean of quality variables, and there was no a major problem in terms of R and Generalized Variance charts that show the variability in the variance structure of the observations. Additionally, the capability analysis showed that the production process was successful in meeting consumer specifications except for only one quality variable.
Key Words: Generalized Variance, Statistical process control, Hotelling T2, Quality control, Process capability analysis, Medium Density Fiberboard |
