Thesis

Computational homogenisation of cracked composite materials using ordinary state-based peridynamics

Downloadable Content

Download PDF
Creator
Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2025
Thesis identifier
  • T17574
Person Identifier (Local)
  • 201762327
Qualification Level
Qualification Name
Department, School or Faculty
Abstract
  • This thesis presents a scalable computational homogenisation framework based on ordinary state-based peridynamics for the effective characterisation of cracked composite materials with periodic microstructures. Conventional continuum methods face challenges in modelling discontinuities such as matrix cracking and fibre-matrix debonding, whereas peridynamics naturally accommodates such damage through its integral formulation. However, integration with computational homogenisation in the ordinary state-based form remains limited. To bridge this gap, a first-order homogenisation scheme is developed within a parallel C++/MPI environment. A key novelty is a raytracing-based bond-breaking algorithm that converts arbitrary cracks into triangle-mesh surfaces, enabling accurate interaction removal and damage tracking. Volumetric periodic boundary conditions are formulated for peridynamic RVEs to ensure displacement periodicity and force anti-periodicity, consistent with classical homogenisation. The framework supports fully 3D scalable simulations and employs harmonic mean sampling for interfacial bond properties to enhance physical fidelity. Numerical studies on fibre-reinforced composite RVEs demonstrate accurate prediction of effective stiffness and local stress fields, validated against analytical and micromechanical models. The results show robust capture of complex crack morphologies and interactions while maintaining parallel efficiency. The proposed framework thus provides a general-purpose and scalable tool for peridynamic homogenisation of damaged composites, forming a foundation for future multiscale and inelastic material modelling.
Advisor / supervisor
  • Oterkus, Selda
  • Oterkus, Erkan
Resource Type
DOI

Relations

Items

Thumbnail Title Date Uploaded Visibility Actions
file details PDF of Thesis T17574 2026-01-29 Public Download