Volume 12, Issue 3, 2003

Detection of Local Preform Defects by Gas Flow Method and Statistical Analysis

Sun K. Kim, Jeremy G. Opperer, Isaac M. Daniel* 

Robert R. McCormick School of Engineering and Applied Science, Center for Intelligent Processing of Composites, Northwestern University, Evanston, IL 60208

*(Author to whom correspondence should be addressed), 

(Received 9/02; accepted 3/03)

ABSTRACT

A method for testing the quality of fibrous preforms using gas flow is proposed. The method detects defects caused while placing the fibre preform in the mould by checking the deviations of measured pressures from the reference expected value at various locations in the mould during steady state gas flow through the enclosed preform. A statistical method is employed to classify the test preform into a normal group or any one of several groups associated with specific types of defects.To implement this method experimentally, a specially designed mould with multiple ports for gas injection and pressure measurement was used. Results obtained for various cases with and without preform defects were compared and analysed.

Keywords: quality control, gas flow method, discriminant analysis, preform defects

© 1999-2001 Advanced Composites Letters. All rights reserved.
No part of the material contained in this site may be copied, reproduced, stored in any form or by any means elsewhere, without the written permission of the editor.