Numerical simulation of elastic wave propagation in textile composite structures
Publication Date
December 31, 2020
Creator
Abstract
This manuscript presents a novel approach allowing damped ultrasonic wave propagation analysis of textile composite structures modelled at a mesoscopic level (i.e. modelling the yarns and matrix distinctively). Current modelling approaches rely on material homogenisation for analysis at a macroscopic scale and thus overlook the effect of textile architecture on wave propagation. This work aims at predicting wave propagation characteristics in damped textile composite structures and the induced complex phenomena for applications in structural health monitoring.
The developed methodology involves mesoscale modelling of a textile composite structure period using a specialised textile modeller for pre-processing as well as conventional finite element methods. This is combined with the periodic structure theory as well as a mode-based reduction method named Craig-Bampton allowing for solving a reduced eigenproblem deriving from the equation of motion. A multiscale approach is used throughout the thesis to enable the comparison of standard wave propagation analysis of composite structures, using homogenised properties, with the more complex analysis proposed in this thesis. The need for this methodology is demonstrated as well as its validity.
The first axis of this thesis describes the methodology for undamped wave propagation analysis in textile composites. Its advantages, such as the prediction of complex phenomena and the possible applications, are thoroughly described and issues discussed. Its increased accuracy over macroscale prediction methods is exposed. A second axis of the thesis is experimental validation of the methodology by means of linear scans of waves measured by a laser vibrometer and generated by a piezoelectric transducer in 3D woven composite samples. It is shown that the numerical mesoscale methodology provides accurate predictions. The third axis is the prediction of dispersion characteristics in large layered assemblies of textile composites. An attempt toward homogenisation of textile composites using a dispersion curves inversion technique based on genetic algorithms is proposed for this purpose. It is concluded that complex textile composites cannot be approximated by simple macroscale models. The last axis of the thesis introduces a damping model to predict the frequency dependent loss factor of waves propagating in these textile composite structures. The strong influence of mesoscale architecture over loss factor is demonstrated.
Item Type
ethesis
Thesis Type
PhD
Supervisors
Subjects (LC)
Associated Schools / Departments
Department of Mechanical, Materials and Manufacturing Engineering (UK)
eprints ID
61490
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
PhDThesis_CorrectedAfterVIVA_VictorThierry.pdf
Type
Full-text
Description
Examined
Size
39.74 MB
Format
Adobe PDF
Checksum (MD5)
b2bc99bad5911b307e3311e642b7d752