Modeling failure of high performance fibers: on the prediction of long-term time-to-failure

Ronald Meester, J Hulshof, Henk Knoester

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

Our objective is to predict the time-to-failure distribution of fibers at loads for which mean time-to-failure is comparable or longer than the fibers’ economic lifetime. We describe load induced time-to-failure of high performance fiber in terms of a classical probabilistic failure model developed by Coleman. Mimicking a series of time-to-failure measurements, using Monte Carlo simulations, we will show how to capture model parameters and their variability using the least squares method and maximum likelihood estimation. It is relatively easy to obtain an accurate prediction for the maximum allowable fiber load such that time-to-failure exceeds a predefined minimum time-to-failure with high probability. However, obtaining a reliable lower prediction limit for time-to-failure at a given, low fiber load is very difficult and requires an unfeasible extensive program of time-to-failure measurements.
Original languageEnglish
Pages (from-to)6277-6290
Number of pages24
JournalJournal of Materials Science
Volume50
Issue number19
Early online date23 Jun 2015
DOIs
Publication statusPublished - Oct 2015

Bibliographical note

PT: J; NR: 24; TC: 0; J9: J MATER SCI; PG: 14; GA: CN1BF; UT: WOS:000358149600004

Fingerprint

Dive into the research topics of 'Modeling failure of high performance fibers: on the prediction of long-term time-to-failure'. Together they form a unique fingerprint.

Cite this