Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Probabilistic method for damage identification in multi-layered composite structures

Kundu, Abhishek, Sikdar, S, Eaton, Mark and Navaratne, R 2018. Probabilistic method for damage identification in multi-layered composite structures. Presented at: 9th European Workshop on Structural Health Monitoring, 10-13 July 2018.

[img]
Preview
PDF - Accepted Post-Print Version
Download (3MB) | Preview

Abstract

The barely visible damages sustained in multi-layered composites can severely jeopardise the structural integrity and operational safety in real-life applications. Traditional intrusive inspections in condition monitoring can signicantly contribute to the cost and time overhead of such operation and is susceptible to errors when relying on manual inspection work ow. Acoustic emission (AE) techniques have received increasing attention in recent years for complex composite structures under service loads. AE is based on detecting acoustic energy emitted from damages sustained in structures (such as fatigue fracture, bre breakage, amongst others). The AE monitoring technique requires solving an inverse problem where the measured signals are linked to the source and nature of damage developed in the structure. However, given the signicant uncertainty around all real-life measurements of structures under operating loads, such as sporadic signals from multiple sources, re ection from boundaries or irregular geometric interfaces and measurement noise, it is essential to explicitly account for these uncertainties in the damage identication algorithms. The current work framework of automated probabilistic damage detection which explicitly models the parameterized uncertainties and conditions them based on measurement data to give probabilistic descriptors of damage metrics. The empirical relationship modelling the AE as a function of damage properties is calibrated with a training dataset. During the online monitoring phase, the spatially correlated time data is utilized in conjunction with the calibrated AE empirical model to infer the probabilistic description of the acoustic emission source within a hierarchical Bayesian inference framework. The methodology is tested on a composite structure consisting of carbon bre panel with stieners subjected to impact and dynamic fatigue loading. The study presents a generalized automated AE- based damage detection methodology which is applicable for structures with dierent geometrical and material properties under conditions of external loading.

Item Type: Conference or Workshop Item (Paper)
Date Type: Completion
Status: Unpublished
Schools: Engineering
Date of First Compliant Deposit: 31 January 2019
Last Modified: 30 Jun 2019 21:51
URI: http://orca.cf.ac.uk/id/eprint/118896

Actions (repository staff only)

Edit Item Edit Item

Downloads

Downloads per month over past year

View more statistics