Automated manufacture of composites: Handling, measurement of properties and lay-up simulations

Prasad Potluri, John Atkinson

    Research output: Contribution to journalArticlepeer-review

    Abstract

    This paper presents some issues relevant to the automated manufacture of low-cost composites. The process of cutting fabric panels from broadcloth and assembling individual pieces had its origins in the clothing industry. The paper investigates the advances made, and the problems encountered in clothing automation projects during the 1980s, with a view to adopting some of the relevant techniques for composites manufacture. Laminate materials, either in the form of prepregs or dry fabrics, exhibit highly non-linear and large deformation behaviour. Hence, a key step towards full-scale automation is the accurate prediction of handling and subsequent processing behaviour of the preforms, with the aid of numerical simulations. The paper describes some automated test methods for measuring the non-linear material properties, required as constitutive properties for simulations. Because of the variety and variability of the preform mechanical properties, rapid and accurate measurements are necessary. A commercial robot has been adopted to conduct a series of tests on dry laminate materials. The tests include intra-ply shear, transverse compression, tension, bending and inter-ply shear. Process simulations such as laying-up, fabric folding and draping have been conducted, based on the measured properties. © 2003 Elsevier Science Ltd. All rights reserved.
    Original languageEnglish
    Pages (from-to)493-501
    Number of pages8
    JournalComposites Part A: Applied Science and Manufacturing
    Volume34
    Issue number6
    DOIs
    Publication statusPublished - Jun 2003

    Keywords

    • A. Fabrics
    • D. Mechanical testing
    • E. Automation
    • E. Lay-up (automated))

    Fingerprint

    Dive into the research topics of 'Automated manufacture of composites: Handling, measurement of properties and lay-up simulations'. Together they form a unique fingerprint.

    Cite this