Aristotle University Of Thessaloniki (AUTH)
Random field modeling of the mechanical properties of composite materials based on their microstructure
Abstract
dc:descriptionThis dissertation investigates the simulation of mechanical properties in composite materials with random microstructure using random fields, addressing the limitations of deterministic methods in capturing stochastic behavior. These random fields serve as input for the response variability analysis of composite structures in the context of the stochastic finite element method. Through high-resolution microstructure imaging techniques, such as SEM and CT, the spatial variability in mechanical properties, such as the Young’s modulus, is investigated. By deriving random fields directly from these images, the proposed framework ensures accurate modeling that reflects real microstructures rather than relying on arbitrary statistical distributions. Homogenization techniques are used in conjunction with the moving window technique to compute mesoscale random fields, which are then employed for conducting macroscopic response analysis. Research work performed during this PhD can be summarized in four key contributions. Firstly, the response variability of composite structures is investigated, whose mechanical properties are described by microstructure-derived random fields with uncertain parameters. Additionally, the mechanical properties of graphene nanoplatelets (GNPs) containing random defects are investigated and a detailed parametric investigation on defect-induced uncertainty is performed. Moreover, a homogenization approach that computes effective bending stiffness properties is implemented, which is further employed for the computation of random fields of bending properties based on real CT-image data of a short fiber composite. Lastly, an efficient random field computation approach is proposed, taking advantage of convolutional neural networks (CNNs) to make nearly instant random field predictions based on image data, and is shown to be applicable to both simulated and real composite microstructures. In this thesis, the microstructure-derived random field approach was examined in detail and it was revealed how it can be employed for investigating the propagation of uncertainty from microstructure to macroscopic response. By considering several types of materials and introducing machine learning tools, such as CNNs, this research offers significant contributions to the analysis and design of heterogeneous materials.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Gavallas, Panagiotis
- Γαβαλλάς, Παναγιώτης
Subjects
dc:subject × 14Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/58060