Multiscale Statistical Process Control of Paper Surface Profiles

This paper presents an approach for setting up a Multiscale SPC procedure that monitors simultaneously two key quality surface phenomena that develop at different scales: roughness and waviness.

Abstract

Paper surface plays a key role in paper quality. Accurate paper surface profiles contain the fundamental raw information of the surface for a wide range of length-scales, to which different aspects of the paper quality are connected. With the goal of exploring the availability of such paper surface data obtained through a mechanical stylus profilometer, we present in this paper an approach for setting up a Multiscale SPC procedure that monitors simultaneously two key quality surface phenomena that develop at different scales: roughness and waviness. The raw profiles, after adequate processing using a multiscale framework based on wavelets, give rise to quantities that can be effectively used to monitor these two phenomena in a simple and integrated way, and therefore be implemented in practice for quality control purposes. The effectiveness of the proposed procedure is assessed by simulation as well as through a pilot study involving real paper surface profiles.