Abstract
This thesis investigates the development and optimisation of modular filament-wound composite joints for hierarchical space frames, focusing on their application in Wrapped Tow Reinforced (WrapToR) truss beam structures. The study aims to address critical challenges associated with the fabrication and connection of lightweight composite truss structures, offering innovative solutions to improve structural efficiency, manufacturability, and scalability.Composite truss structures are highly valued for their superior strength-to-weight ratios, making them particularly suitable for aerospace, automotive, and architectural applications. However, their practical use has been limited by challenges related to fabricating and joining complex geometries. Traditional joining methods, such as mechanical fastening and adhesive bonding, often compromise structural integrity due to stress concentrations and the anisotropic properties of composite materials. Recent advancements in filament-wound trusses, such as the WrapToR truss beams, have demonstrated significant potential by combining mechanical performance with design flexibility. Despite these advances, a lack of robust, scalable joint solutions remains a critical barrier to wider adoption.
The primary objective of this research is to develop a modular filament-wound joint concept capable of connecting multiple truss beams to form hierarchical space frame structures. The research focuses on optimising joint designs to achieve structural efficiency while maintaining manufacturability and validating these designs through numerical simulations and experimental testing.
To address these objectives, this study adopts a multi-disciplinary methodology. The development of coaxial joints represents the foundational phase, where WrapToR trusses with varying cross-sectional sizes are connected. Finite element analysis (FEA) models were employed to predict the mechanical performance of these joints, achieving high accuracy with deviations of less than 4.5\% compared to experimental results. These findings validated the feasibility of the proposed designs and informed subsequent development.
A key innovation of this research is the creation of a graph-theory-based algorithm for three-dimensional winding path planning. This algorithm addresses critical manufacturing constraints, such as avoiding filament overlaps and accommodating geometric complexities, ensuring that winding paths are continuous and manufacturable. The algorithm was further validated through the production of complex T-joint and cross-joint samples, demonstrating its applicability to real-world manufacturing scenarios.
Building on this foundation, the study introduced an optimisation framework for T-joint designs. A genetic algorithm was employed to minimise both the mass and deformation of the joints under specific loading conditions. To ensure practicality, the algorithm integrated manufacturability constraints, balancing material efficiency with structural performance. The optimisation results revealed critical members within the T-joint, allowing for targeted material distribution along primary load paths while identifying peripheral members that contributed minimally to structural integrity.
Experimental trials confirmed the effectiveness of an optimised T-Joint design, with the production of high-quality joint samples. While unnoticed manufacturing defects led to lower than expected rigidity in the jointed structure due to some members essentially being pre-buckled, incorporating the estimated impact of these defects into the FE model allowed for accurate prediction of stiffness.
The findings of this research significantly advance the understanding of modular filament-wound joints. By integrating structural optimisation with manufacturing constraints, the study demonstrates how scalable and efficient connections can be achieved for complex composite truss systems. The development of the graph-based path planning algorithm and the genetic optimisation framework addresses the dual challenges of manufacturability and performance, paving the way for broader applications of hierarchical space frames in engineering.
This research provides a foundation for the further advancement of composite truss systems, with implications for various engineering fields requiring lightweight, adaptive structures. Future work will focus on automating the winding process, refining the optimisation framework for additional joint configurations, and exploring new manufacturing techniques to reduce imperfections. These developments will enhance the scalability and reliability of composite truss systems, further unlocking their potential in high-performance engineering applications.
| Date of Award | 30 Sept 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Terence Macquart (Supervisor) & Ben K S Woods (Supervisor) |