Multi-scale modelling and material characterisation of textile composites for aerospace applications

Pan, Qing (2016) Multi-scale modelling and material characterisation of textile composites for aerospace applications. PhD thesis, University of Nottingham.

PDF (Corrected PDF) (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (18MB) | Preview


Textile composites offer an excellent alternative to metallic alloys in the aerospace engineering due to their high specific stiffness and strength, superb fatigue strength, excellent corrosion resistance and dimensional stability. In order to successfully apply these materials to engineering problems, a methodology to characterise and predict the constitutive response of these materials is essential. The lack of the modelling tools for modern textile composites that would facilitate systematic analysis and characterisation of these materials hinders the wide adoption of such material systems in engineering applications. This defines the focus of the project as represented in this thesis.

A multi-scale modelling methodology has been established for the material characterisation and representing the constitutive response of the material at a macroscale. For material characterisation at micro- and mesoscales, an automated material characterisation toolbox, UnitCells©, has been employed and substantially developed in both the scope and complexity through this project.When applying this toolbox, the user selects the required type of a textile or unidirectional reinforcements and provides a parametric input, based on which a finite element model of a unit cell for the composites is generated automatically. The effective properties that can be predicted using this toolbox include stiffness, thermal expansion coefficient, thermal and electric conductivities, static strength and dynamic strength (associated with deformation localisation as the limit of the applicability of unit cells but a conservative estimate of the material strength). There are seven types of microscale models and eleven types of mesoscale models available in the toolbox at present.

To represent a constitutive relationship for textile composites at a macroscale, the artificial neural network (ANN) algorithm has been adapted and developed into a useful modelling tool, referred to as the ANN system. A criterion defining an ultimate failure of the material has been proposed. The outcome has made it possible for a user defined material subroutine to be established which can be employed in the analysis of structures made of such textile composites by providing the effective constitutive behaviour of them in a most efficient manner. As a validation, ANN system was used to predict the critical velocities of three types of layer-to-layer interlock 3D woven composite panel subject to ballistic loading. The predicted results matched well with the testing results. Furthermore, as an illustration of potential capability, the ANN system has been used to simulate impact of a textile composite fan blade containment casing in an idealised fan blade off scenario.

Through the project, the capability of predicting the impact behaviour of textile composites has been established. This involves unit cell modelling at micro-/mesoscales for material characterisation, strength prediction with due consideration of strain rate sensitivity of the constituent materials, and ANN system to deliver the characterised constitutive relationship in terms of a user defined material subroutine for practical applications at macroscale, such as structural impact analysis.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Li, Shuguang
Long, Andrew C.
Sitnikova, Elena
Keywords: Textile Composties, Multi-scale Modelling, Artificial Neural Network, Aerospace Applications
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Faculties/Schools: UK Campuses > Faculty of Engineering
Item ID: 33396
Depositing User: Pan, Qing
Date Deposited: 16 Aug 2016 10:20
Last Modified: 17 Dec 2017 01:55

Actions (Archive Staff Only)

Edit View Edit View