Cell Configurations in the ANOVA Interaction Model to Identify Two-Dimensional Patterns in Reel Data, International Control Systems Conference (TAPPICon26)

The interaction model of Reel Statistics has been in existence for decades. The seminal paper on Analysis of Variance published by John Burns in 1973 identified advanced statistical features that the QCS industry was not prepared to use and implement. In part this lack of acceptance was due to the limitations of early computers. Some QCS suppliers needed to develop custom models to estimate basic Reel Statistics for a limited number of databoxes and scans. Another reason for the lack of acceptance of the interaction model was that the dimensions of the interaction (i.e. cell size) were not established in the model, leaving that selection to the user. Hence, its dimensions could vary between users and cause different results. Unlike the typical variance partition analysis (VPA), the interaction model divides variance into four components (Machine Direction (MD), Cross Direction (CD), Interaction (INT) and Residual (RES)) and uses F-Tests to determine when a component is not statistically significant. This interaction model’s INT component identifies the presence of 2-dimensional (2D) patterns that cannot be attributed to MD, CD, or random variations (RES). These 2D patterns include stock skating, scanner stepping, CD control instability or asynchronous MD aliasing (not synchronized to scan time). This paper uses Excel to create distinct 2D patterns in the simulated Reel Data and then investigates how the Interaction can be structured to confirm the significance of that pattern. The simulation allows the experimenter to explore the type of 2D patterns and their magnitude. Some actual Reel Data is also presented.

The goal of this paper is to investigate potential cell configurations for the interaction and what 2D patterns can best be revealed with each configuration. This knowledge can lead to using the interaction model to confirm the existence of 2D patterns in the Reel Data that are not pure MD or CD and thereby work as a troubleshooting tool to help identify the underlying cause of variability that is often wrongly classified as random.

TAPPI conference proceedings and presentations, technical papers, and publication articles provide technical and management data and solutions on topics covering the Pulp, Paper, Tissue, Corrugated Packaging, Flexible Packaging, Nanotechnology and Converting Industries.

Simply select the quantity, add to your cart and your conference paper, presentation or article will be available for immediate download.