Quantification of uncertainty modelling in stochastic analysis of FRP composites

Srinivas Sriramula, Marios K. Chryssanthopoulos

Research output: Contribution to journalArticle

166 Citations (Scopus)

Abstract

The extensive use of FRP composite materials in a wide range of industries, and their inherent variability, has prompted many researchers to assess their performance from a probabilistic perspective. This paper attempts to quantify the uncertainty in FRP composites and to summarise the different stochastic modelling approaches suggested in the literature. Researchers have considered uncertainties starting at a constituent (fibre/matrix) level, at the ply level or at a coupon or component level. The constituent based approach could be further classified as a random variable based stochastic computational mechanics approach (whose usage is comparatively limited due to complex test data requirements and possible uncertainty propagation errors) and the more widely used morphology based random composite modelling which has been recommended for exploring local damage and failure characteristics. The ply level analysis using either stiffness/strength or fracture mechanics based models is suggested when the ply characteristics influence the composite properties significantly, or as a way to check the propagation of uncertainties across length scales. On the other hand, a coupon or component level based uncertainty modelling is suggested when global response characteristics govern the design objectives. Though relatively unexplored, appropriate cross-fertilisation between these approaches in a multi-scale modelling framework seems to be a promising avenue for stochastic analysis of composite structures. It is hoped that this review paper could facilitate and strengthen this process. (C) 2009 Elsevier Ltd. All rights reserved.

Original languageEnglish
Pages (from-to)1673-1684
Number of pages12
JournalComposites Part A: Applied Science and Manufacturing
Volume40
Issue number11
DOIs
Publication statusPublished - Nov 2009

Keywords

  • polymer-matrix composites (PMCs)
  • mechanical properties
  • micro-mechanics
  • statistical properties/methods
  • representative volume element
  • fiber-reinforced composites
  • polymer matrix composites
  • moving-window technique
  • heterogeneous materials
  • laminated composites
  • mechanical-properties
  • elastic composites
  • failure analysis
  • unidirectional composites

Cite this

Quantification of uncertainty modelling in stochastic analysis of FRP composites. / Sriramula, Srinivas; Chryssanthopoulos, Marios K.

In: Composites Part A: Applied Science and Manufacturing, Vol. 40, No. 11, 11.2009, p. 1673-1684.

Research output: Contribution to journalArticle

@article{9428cfb1ca00414aa881a2e30596c486,
title = "Quantification of uncertainty modelling in stochastic analysis of FRP composites",
abstract = "The extensive use of FRP composite materials in a wide range of industries, and their inherent variability, has prompted many researchers to assess their performance from a probabilistic perspective. This paper attempts to quantify the uncertainty in FRP composites and to summarise the different stochastic modelling approaches suggested in the literature. Researchers have considered uncertainties starting at a constituent (fibre/matrix) level, at the ply level or at a coupon or component level. The constituent based approach could be further classified as a random variable based stochastic computational mechanics approach (whose usage is comparatively limited due to complex test data requirements and possible uncertainty propagation errors) and the more widely used morphology based random composite modelling which has been recommended for exploring local damage and failure characteristics. The ply level analysis using either stiffness/strength or fracture mechanics based models is suggested when the ply characteristics influence the composite properties significantly, or as a way to check the propagation of uncertainties across length scales. On the other hand, a coupon or component level based uncertainty modelling is suggested when global response characteristics govern the design objectives. Though relatively unexplored, appropriate cross-fertilisation between these approaches in a multi-scale modelling framework seems to be a promising avenue for stochastic analysis of composite structures. It is hoped that this review paper could facilitate and strengthen this process. (C) 2009 Elsevier Ltd. All rights reserved.",
keywords = "polymer-matrix composites (PMCs), mechanical properties, micro-mechanics, statistical properties/methods, representative volume element, fiber-reinforced composites, polymer matrix composites, moving-window technique, heterogeneous materials, laminated composites, mechanical-properties, elastic composites, failure analysis, unidirectional composites",
author = "Srinivas Sriramula and Chryssanthopoulos, {Marios K.}",
year = "2009",
month = "11",
doi = "10.1016/j.compositesa.2009.08.020",
language = "English",
volume = "40",
pages = "1673--1684",
journal = "Composites Part A: Applied Science and Manufacturing",
issn = "1359-835X",
publisher = "Elsevier Limited",
number = "11",

}

TY - JOUR

T1 - Quantification of uncertainty modelling in stochastic analysis of FRP composites

AU - Sriramula, Srinivas

AU - Chryssanthopoulos, Marios K.

PY - 2009/11

Y1 - 2009/11

N2 - The extensive use of FRP composite materials in a wide range of industries, and their inherent variability, has prompted many researchers to assess their performance from a probabilistic perspective. This paper attempts to quantify the uncertainty in FRP composites and to summarise the different stochastic modelling approaches suggested in the literature. Researchers have considered uncertainties starting at a constituent (fibre/matrix) level, at the ply level or at a coupon or component level. The constituent based approach could be further classified as a random variable based stochastic computational mechanics approach (whose usage is comparatively limited due to complex test data requirements and possible uncertainty propagation errors) and the more widely used morphology based random composite modelling which has been recommended for exploring local damage and failure characteristics. The ply level analysis using either stiffness/strength or fracture mechanics based models is suggested when the ply characteristics influence the composite properties significantly, or as a way to check the propagation of uncertainties across length scales. On the other hand, a coupon or component level based uncertainty modelling is suggested when global response characteristics govern the design objectives. Though relatively unexplored, appropriate cross-fertilisation between these approaches in a multi-scale modelling framework seems to be a promising avenue for stochastic analysis of composite structures. It is hoped that this review paper could facilitate and strengthen this process. (C) 2009 Elsevier Ltd. All rights reserved.

AB - The extensive use of FRP composite materials in a wide range of industries, and their inherent variability, has prompted many researchers to assess their performance from a probabilistic perspective. This paper attempts to quantify the uncertainty in FRP composites and to summarise the different stochastic modelling approaches suggested in the literature. Researchers have considered uncertainties starting at a constituent (fibre/matrix) level, at the ply level or at a coupon or component level. The constituent based approach could be further classified as a random variable based stochastic computational mechanics approach (whose usage is comparatively limited due to complex test data requirements and possible uncertainty propagation errors) and the more widely used morphology based random composite modelling which has been recommended for exploring local damage and failure characteristics. The ply level analysis using either stiffness/strength or fracture mechanics based models is suggested when the ply characteristics influence the composite properties significantly, or as a way to check the propagation of uncertainties across length scales. On the other hand, a coupon or component level based uncertainty modelling is suggested when global response characteristics govern the design objectives. Though relatively unexplored, appropriate cross-fertilisation between these approaches in a multi-scale modelling framework seems to be a promising avenue for stochastic analysis of composite structures. It is hoped that this review paper could facilitate and strengthen this process. (C) 2009 Elsevier Ltd. All rights reserved.

KW - polymer-matrix composites (PMCs)

KW - mechanical properties

KW - micro-mechanics

KW - statistical properties/methods

KW - representative volume element

KW - fiber-reinforced composites

KW - polymer matrix composites

KW - moving-window technique

KW - heterogeneous materials

KW - laminated composites

KW - mechanical-properties

KW - elastic composites

KW - failure analysis

KW - unidirectional composites

U2 - 10.1016/j.compositesa.2009.08.020

DO - 10.1016/j.compositesa.2009.08.020

M3 - Article

VL - 40

SP - 1673

EP - 1684

JO - Composites Part A: Applied Science and Manufacturing

JF - Composites Part A: Applied Science and Manufacturing

SN - 1359-835X

IS - 11

ER -