Design of a Dynamic Quality Control System for Textile Processes

Abstract

An attempt has been made to apply the structural equations published during the last 60 years for designing a dynamic quality control system for dry and wet textile processes that are either continuous or contiguous and are serially connected with time lags.

This new system provides process averages and control limits that are relative to the conditions of the prior processes. According to the new system the changes observed in the prior process will update the process averages and control limits of the current process using the structural relationship between the two stages. By obtaining more accurate control limits, the root causes of the out of control situations will be determined precisely, and unnecessary corrective actions that are detrimental to quality monitoring improvement be minimized.

The major research task was to identify all published papers and sort out clearly defined input and output parameters that are essential in determining the structural relationships between the various process stages that are serially connected.

The next challenge was to align and consolidate the multiple equations to a single set at each stage in such a way that a dynamic system can be developed by combining all process steps in sequence, linking all input and outputs parameters. In the jth process, the output Yj is expressed as a function of Yj-1 of the previous process and 'm' new input factors zj (zj1, zj2, ..., zjm);

In any given two contiguous processes, the input (z) and output (Y) relationships are more than one in most cases. Using FAMSE technique, the multiple structural equations are to be consolidated into a single structural equation. Using Variance Tolerancing and Channeling technique the output variances are computed and thus the dynamic process averages and their relative dynamic control limits are calculated.