Files in this item



application/pdf8803023.pdf (7MB)Restricted to U of Illinois
(no description provided)PDF


Title:A Fault Classification System for Quality and Productivity Improvement in Continuous Manufacturing Processes
Author(s):Dooley, Kevin John
Doctoral Committee Chair(s):Kapoor, Shiv G.
Department / Program:Mechanical Engineering
Discipline:Mechanical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Engineering, Mechanical
Abstract:The manufacturing industry has recognized the need to improve the quality and productivity of its products and processes in order to increase its competitive position. In order to establish a high quality product companies install "quality systems" to improve product and process quality through intelligent manipulation and design of processes and resources over time. The design of quality control (QC) windows is a major step in the operation of an integrated quality system. A QC window covers a portion of the process where elements of the process transfer function are monitored and evaluated over time.
The work presented here encompasses the evaluation element of a QC window. The Fault Classification System greatly enhances the chance of quality improvement by detecting and classifying faults as they occur. Specifically, a continuous process is modeled by time series and three residual analysis techniques--the Chi-Square test, the cusum and the autocorrelation chart--are used to identify changes in the common cause variability of the process. A rule base classifies faults as a shift in the process mean, variance, or transfer function parameters. The system then estimates fault magnitude and time of occurrence via a least squares scheme. This additional diagnostic information about the fault occurrence makes the task of fault diagnosis simpler and more informative.
An end milling process is analyzed via a physical experiment to verify the FCS performance. Forces from the end milling operation are monitored over time and modeled by time series. Changes in the cutting tool feedrate are introduced and lead to corresponding changes in average forces observed. The FCS detects the shift in the force signal and estimates when the feedrate shift occurred. This knowledge helps in identification of the root cause of the feedrate shift, as well as being useful for direct feedback control.
Issue Date:1987
Description:231 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1987.
Other Identifier(s):(UMI)AAI8803023
Date Available in IDEALS:2014-12-15
Date Deposited:1987

This item appears in the following Collection(s)

Item Statistics