RMIT University
Browse

Modelling the tension and compression strengths of polymer laminates in fire

journal contribution
posted on 2024-11-01, 04:30 authored by Adrian Mouritz, Zenka Mathys, Arthur Gibson
Thermo-mechanical models are presented for predicting the time-to-failure of polymer laminates loaded in tension or compression and exposed to one-sided radiant heating by fire. Time-to-failure is defined as the time duration that a polymer laminate can support an externally applied load in a fire without failing. The models predict the temperature rise and through-thickness temperature profile in a hot decomposing laminate exposed to fire. Using this thermal data, mechanics-based models based on residual strength analysis are used to calculate the time-to-failure. A preliminary evaluation of the accuracy of the models is presented using failure times measured in fire-under-load tests on a woven glass/vinyl ester laminate. The model was evaluated at temperatures between -250 and 800 °C by testing the laminate at heat flux levels between 10 and 75 kW/m2. It was found that the time-to-failure of the laminate decreased with increasing heat flux and increasing applied stress for both the compression and tension load conditions. The tests also revealed that the failure times were much shorter (by about one order of magnitude) when the laminate was loaded in compression. The models can predict the time-to-failure with good accuracy for both compression and tension loading for certain heat flux levels. However, because the models have only been evaluated for one type of laminate (woven glass/vinyl ester), further evaluation is necessary for other laminate systems. The paper also presents new experimental insights into the strengthening mechanisms of laminates at high temperature.

History

Related Materials

  1. 1.
    DOI - Is published in 10.1016/j.compscitech.2006.07.038
  2. 2.
    ISSN - Is published in 02663538

Journal

Composites Science and Technology

Volume

67

Start page

551

End page

564

Total pages

14

Publisher

Elsevier Science

Place published

Oxford

Language

English

Former Identifier

2006005849

Esploro creation date

2020-06-22

Fedora creation date

2010-12-22

Usage metrics

    Scholarly Works

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC