A study of surface fracture in paste extrusion using signal processing

BD Russell, J Lasenby, Stuart Blackburn, DI Wilson

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Gross circumferential fracture (GCF) is a surface defect sometimes observed during the extrusion of particulate pastes. In this study particular attention was paid to the frequency of fracture in pastes undergoing ram extrusion through dies with circular, square and rectangular slot cross section. Three methods for estimating the fracture frequency were employed. The first was determined from the observed fracture periodicity, the second was determined from the application of short time Fourier transformation (STFT) techniques to the pressure signal measured at the die plate and the third was the application of the geometrical result first reported by Domanti and Bridgwater (2000). All were found to be in agreement within experimental error where failure was observed. In some materials the fracture frequency was still observed in the frequency domain after fracture ceased in the extrudate, indicating the existence of a flow instability in the paste: if the extrudate was smooth throughout the process STFT detected no peak. Fracture was more pronounced with short dies and/or where the paste experienced high wall shear stress. Square dies were more prone to fracture than circular dies (corner effects) and the fracture spacing could be related directly to the hydraulic diameter. Rectangular slot dies were more prone to irregular fracture. It was shown that lowering the die wall friction by using Teflon-lined dies reduced the propensity of the paste to fracture for all die geometries tested. (c) 2006 Springer Science + Business Media, Inc.
Original languageEnglish
Pages (from-to)2895-2906
Number of pages12
JournalJournal of Materials Science
Volume41
Early online date21 Apr 2006
DOIs
Publication statusPublished - 21 Apr 2006

Fingerprint

Dive into the research topics of 'A study of surface fracture in paste extrusion using signal processing'. Together they form a unique fingerprint.

Cite this