PhD projects 2010 cohort
- Unified nonlinear damage model for fatigue delamination onset and growth in fibre reinforced plastics
- On the relationship between layup time, material properties and mould geometry in the hand layup process
- Buckling of stiffened variable stiffness panels
- The future of sheet prepreg layup
- Pseudo-ductility of thin ply angle-ply laminates
- Simple tools for impact assessment
- On delamination migration in composite laminates
- Novel self-healing systems: expanding and inhibited healing agents
- On the formation of composite materials via ultrasonic assembly
- Towards CNT fibre/polymer composites
- Post-buckling of variable-stiffness shell structures
Unified nonlinear damage model for fatigue delamination onset and growth in fibre reinforced plastics
Student: Jamie Blanchfield
Supervisors: Stephen Hallett and Giuliano Allegri (Imperial College London)
Despite their excellent in-plane strength, composite structures are very weak through the thickness and are therefore prone to delamination failures at the interfaces between plies. Cyclic loading of these structures, such as occurs in aircraft engines, can lead to delamination onset and growth, and the ability to accurately model the evolution of this type of damage is critical to reducing the large costs associated with experimental testing of composites.
This project aims to extend a recently developed unified mode II fatigue damage evolution law so that it can account for mixed mode damage within a consistent thermodynamic framework. In order to do this, material SN curves for damage initiation are required for both mode I and mixed mode conditions at a range of stress ratios. After a new mixed-mode damage law has been postulated, it will be implemented numerically within a commercially available Finite Element code, and validated against experimental damage data from a suitable composite component. If successful, this has the potential to provide a robust engineering tool for the analysis of fatigue damage in composites.
On the relationship between layup time, material properties and mould geometry in the hand layup process
Student: Dominic Bloom
Supervisors: Kevin Potter and Carwyn Ward
Prepreg hand layup seems to have achieved black art status. With attention shifting towards automated processes, lack of repeatability, inconsistent quality and high cost are presented as arguments against the continued use of manual processes. However, the state of the art Automated Tape Laying (ATL) and Automated Fibre Placement (AFP) does not allow for the manufacture of small, complex parts such as secondary structures, due to their relative geometric complexity. These are often made from woven prepreg.
The project will initially focus on gaining an understanding of the hand layup process, with particular focus on material properties, as there is no conclusive data which relates prepreg material properties to performance during hand layup. Effort will be made to identify the extent to which standard material tests (such as ASTM) can be used to identify a material’s suitability for layup, and a new test will be developed to characterise materials in a more relevant manner. Each material property will be considered in turn, and its contribution to overall “usability” examined.
A series of experiments involving laminators will help clarify the limits of both human and material performance, and will help to justify a case for the automation of small complex parts. The knowledge will be used to design automated solutions to hand layup, be it for the entire process, or for the elements which would benefit from automation the most, being mindful of the areas in which human capability remains unsurpassed.
Buckling of stiffened variable stiffness panels
Student: Broderick Coburn
Supervisors: Paul Weaver and Jacopo Ciambella (Sapienza Università di Roma)
Over the past five decades the aerospace industry has seen the increased use and development of composite materials and structures, largely due to their superior specific strength and stiffness in comparison with metals. These two major attributes, coupled with their increased tailorability, enable the design of lighter, more efficient structures. Traditional tailoring is limited to modifying the orientation of laminate plies to achieve the desired performance. However, recent advancements in tape laying and fibre placement technologies, have led to the possibility of laminates with varying fibre orientations in the plane of a ply, thereby creating variable stiffness (VS) structures.
These VS laminates greatly expand the design space and provide designers with additional degrees of freedom and tailorability. In recent years, analytical methods and finite element analysis (FEA) have shown that VS laminates can deliver a significant improvement in buckling and post-buckling performance for plates and shells. A potential application, which exploits this enhanced buckling performance, is the use of VS laminates as the skin of a stiffened panel. Here, the VS skin is designed to redistribute in-plane loads towards the stiffeners, which act as panel breakers, providing an expected increase in buckling performance, thereby facilitating the design of lighter and more efficient structures.
The present work was conducted with the main aim of developing fast, robust and accurate tools for the linear buckling analysis of novel stiffened VS panels. The Ritz energy method was used to develop semi-analytical models for various plate configurations subject to uniaxial and biaxial compression; firstly, plates exhibiting discontinuously varying stiffness terms with bend-stretch coupling, a feature commonly present in the skin-flange region of stiffened panels; secondly, thick plates and sandwich panels with continuously varying stiffness terms, through fibre-steering, by including a first-order shear deformation theory; and thirdly, blade stiffened VS panels by utilising the previously developed methods to capture the important features of discontinuous and continuous stiffness variations and transverse shear flexibility. Benchmarked against a commercial FEA package, the semi-analytical models are shown to be both an accurate and computationally efficient alternative, and thus well suited for design and optimisation.
To conclude, design and heuristic optimisation case studies for blade stiffened panels, with both straight fibre and VS skins, were performed applying practical design and failure constraints. Optimal VS designs showed improvements in structural efficiency compared to standard configurations, with weight savings over 5% and 20% for VS skins satisfying and neglecting the 10% rule, applied only for 0o and 90o plies, respectively. Furthermore, the mass reductions were shown to be achievable utilising relatively few VS plies in some cases, thus providing an avenue for application without the need for significant changes in stiffener or skin design, rates of deposition, or design rules and guidelines.
Broderick was awarded a Faculty of Engineering Commendation for his thesis.
The future of sheet prepreg layup
Student: Mike Elkington
Supervisors: Kevin Potter and Carwyn Ward
For many complex composite parts, ‘Hand Layup’ is currently the only way they can be successfully manufactured. Despite its importance to the composites industry, the hand layup process has remained relatively unchanged for nearly 30 years. The overall vision of this project was to first gain a greater understanding of the hand lamination process itself, and then apply this knowledge to improve the process in a variety of ways. An ergonomics study of the lamination process revealed that despite how complex layup was, the actions made by laminators could be broken up into a small collection of techniques. Based on this research, a new approach to layup was developed which produced time savings of up to 40% while dramatically simplifying the process.
This simplification opened the door to apply automation techniques to layup. A two stage process was developed, where individual plies were first shaped in a press to create all the required in-plane deformation. They were then stuck down onto the mould using a new robotic process. A six axis robotic arm was equipped with a series of bespoke end effectors designed to enable layup over a wide variety of complex shapes. This was then used to layup a series of different mould shapes of increasing complexity, building up to the fully automated manufacture of a large sandwich panel style component.
Watch Michael introducing his work in this short YouTube video.
Pseudo-ductility of thin ply angle-ply laminates
Student: Jonathan Fuller
Supervisors: Michael Wisnom and Ian Bond
Angle-ply laminates show promise of high strains to failure, but specimens can fail prematurely due to edge delaminations, before any non-linear behaviour can develop. The energy available to propagate delaminations is proportional to ply thickness and reducing ply blocking has been shown to increase laminate strength. The current approach, using thin plies of carbon fibre reinforced polymer, aims to suppress these delaminations and allow higher strains to be reached before final laminate failure.
Numerical modelling of the laminate behaviour in tension shall be performed using ABAQUS and MATLAB, in order to provide an optimised layup that exhibits high strain to failure while maintaining a high stiffness and strength. These laminates will be mechanically tested to validate the models. It is expected that axial strains in excess of 10% can be demonstrated through damage suppression and fibre reorientation.
Following selection of an optimal ply angle, the project will investigate a multi-directional case using non-straight, or ‘wavy’ fibre tapes. It is believed that high strains to failure will be shown in tension due to a multitude of fibre reorientations along the length of the material. Laminates consisting of various combinations of wavy and straight fibre plies shall be modelled and tested with the aforementioned aim of high strains and maintenance of strength and stiffness. Opportunities may also exist for the incorporation of a ductile polymer in to the fibre tapes, to possibly further improve the strain to failure behaviour.
Simple tools for impact assessment
Student: Salah Muflahi
Supervisors: Stephen Hallett, Giuliano Allegri (Imperial College London), Galal Mohamed
Current high fidelity finite element analysis tools for impact assessment can be extremely complex and have large run times which can slow the design process. The main aims of this project are to create much simpler tools for the assessment of composite plates under impact and verify these against existing models and experimental results. As an early stage design tool, this can be used to save time and verify whether a given configuration is feasible.
Finite-element models are created in LS-DYNA, and by looking at the optimum solution to sufficiently capture the mechanisms for damage, whilst reducing the computational cost by using Cohesive Zone Modelling (CZM) techniques with shell elements. Part of this work is investigating the influence of the number of interfaces modelled, to see whether using a single interface can be used to sufficiently model the initiation and propagation of damage in a laminated structure for design purposes.
Initial work has been carried out to see whether a closed-form analytical model could be used, however due to the complexities in the model for a CFFF (Cantilever) plate the simplified finite-element method was chosen as the preferred solution.
On delamination migration in composite laminates
Student: Francesca Pernice
Supervisors: Stephen Hallett, Giuliano Allegri (Imperial College London) and Luiz Kawashita
This project is investigating the migration of cracks through the thickness of laminated composite materials. Failure of laminated composites is a complex process, involving a number of different mechanisms. Damage can occur in the fibres, in the matrix or at the interface between layers (delamination). The failure mechanism becomes even more complex since the different types of damage interact with each other to cause overall failure. Crack migration through the thickness occurs due to the interaction between cracks in the matrix and delamination. When these two types of flaws meet in a laminate, the delamination can migrate from one interface to the next. The process allows the delamination to find and move to the weakest interface, finally causing the complete failure of a structure.
This damage mechanism is an important cause of final failure of laminated composite structures. Nevertheless, it is usually neglected in finite element analyses, due to the difficulty in modelling the interaction between the different damage mechanisms involved using the currently available finite element tools.
This project’s main focus is on crack migration through the thickness in laminated composites. Experimental tests will be carried out to gain a better understanding of the physics of the processes involved. Test cases will be developed in order to examine the damage progression and measure the quantities involved (e.g. fracture energy). Finite element methodologies able to predict the crack interaction will be developed and their results validated against experiments. Numerical models will help investigate the failure process and predict the damage progression in laminate composite structures.
Novel self-healing systems: expanding and inhibited healing agents
Student: Steven Rae
Supervisors: Ian Bond, Richard Trask and Duncan Wass
The concept of self-healing materials has emerged from the reticence that exists in composite design, especially in aerospace structures. This concern emanates from composite materials’ poor interlaminer properties and therefore tendency to perform badly when subject to impact events, typically manifesting as matrix cracking, delamination, and fibre debonding. With even microscopic damage having the potential to grow under fatigue loading until the structure’s mechanical properties are diminished, composite structures are manufactured with high built in safety factors and structural redundancy to counteract inevitable defect creation.
By developing self-healing materials, these defects can be addressed before (or after) they are allowed to grow, thus reducing the requirement for structural redundancy and capitalise on the mass savings that result. The chemistry behind healing mechanisms, and methods of incorporating healing functionality itself, has been intensely researched by many groups in recent years. Whilst impressive results have been observed, and respecting the advancements that have been achieved, there still exist challenges which need to be addressed to allow for effective and fully autonomous self-healing systems.
Many studies report thermal activation of polymerisation reactions, pre-mixing of healing agents, manual closing of crack planes to increase the relative volume of healing agent, or artificial opening of crack planes to increase infiltration and alleviate tensile stresses on the healing agent. Fundamentally however, achieving high healing efficiencies relies on delivering an adequate volume of healing agent(s) in a stoichiometric ratio and achieve effective mixing, or relies on exposing embedded catalyst to drive polymerisation. We aim to address some of these challenges and reduce the dependency on external stimulus to increase healing efficiency in an autonomous manner through two different approaches. Firstly, the problems associated with incorporation of catalyst into the matrix, achieving stoichiometric ratios, and effective mixing, can be addressed using a single part healing chemistry that requires no additional stimulus or catalyst after release to polymerise. We have therefore investigated a potential route to ‘inhibited healing’ whereby a resin is actively prevented from undergoing polymerisation until released from the delivery vessel, whereupon polymerisation occurs rapidly and autonomously. Secondly, problems associated with mixing, reducing fibre disruption from vascule incorporation, delivering adequate volume from smaller reserves, achieving high proportions of infiltration, or to address larger damage voids and bridge wider separations, can be achieved by creation of volume in the healing agent itself. We have investigated different chemical systems to produce a structural polymer with a volume greater than the sum of its constituent parts, explored methods of tailoring its chemical and structural properties, and assessed its ability to repair not only the relatively small volumes associated with damage within laminate structures, but also the larger damage volumes associated with impacted sandwich structures.
On the formation of composite materials via ultrasonic assembly
Student: Marc Scholz
Supervisors: Richard Trask and Bruce Drinkwater
Acoustic levitation techniques have been widely studied within the biological and medical disciplines, primarily to manipulate cells and molecules whose size range lies outside the capabilities of optical tweezers. Since, ultrasonic assembly has been applied more widely, with the trapping of micron- to millimetre-size objects of different shapes and sizes, and the formation of ordered arrays of particles having become possible. The research presented in this thesis investigates the feasibility of applying ultrasonic particle manipulation methodologies to manufacture short fibre reinforced polymer composites.
A series of ultrasonic devices is developed allowing the manufacture of thin layers of anisotropic composite material. Strands of unidirectional reinforcement are, in response to the acoustic radiation force, shown to form inside various matrix media. The technique proves suitable for both photo-initiator and temperature controlled polymerisation mechanisms.
To further explore key parameters in the design of ultrasonic devices, a number of linear acoustic models are developed. One- and two-dimensional finite element analysis are employed to study the resonance characteristics, compute the acoustic pressure, and calculate the acoustophoretic force on small spherical particles. A range of fibre architectures that can be generated with devices of up to eight transducer elements is explored by plane and spherical wave propagation methods.
A separate study analyses the dynamic response of both an elastic sphere and cylinder placed in a standing wave field by solving the equations of non-linear fluid dynamics for arbitrary angles of radiation incidence. A comparison with analytical results shows good agreement in the limit of small particles. For large particles, the acoustic radiation force is further evaluated across a range of pressure amplitudes, and for a number of initial particle positions.
Finally, a series of glass fibre reinforced composite samples constructed via the ultrasonic assembly process are subjected to tensile loading and the stress-strain response is characterised. Structural anisotropy is clearly demonstrated, together with a 43% difference in failure stress between principal directions. The average stiffnesses of samples strained along the direction of fibre reinforcement and transversely across it were 17.66±0.63MPa and 16.36±0.48MPa, respectively.
Towards CNT fibre/polymer composites
Student: James Trevarthen
Supervisors: Sameer Rahatekar, Michael Wisnom and Krzysztof Koziol (University of Cambridge)
The excellent nano-scale mechanical properties of carbon nanotubes (CNTs) are well documented, but remain to be effectively exploited in macro-scale materials. Recently, CNT fibres spun from sheets of aligned CNT have shown strengths and stiffnesses to rival other high-performance fibres as well as ductility and high strains-to-failure after simple chemical treatment.
The aim of this project is to produce high-performance, ductile nanocomposites using CNT fibres. This will be achieved by, firstly, understanding the complex fibre/matrix interface and then using this knowledge to produce a composite system with the optimum combination of fibre, matrix and interface to achieve high-performance. If successful, this will overcome the significant limitation of brittleness in conventional-fibre reinforced composites as well as demonstrate a route to successful exploitation of nano-scale CNT properties.
Post-buckling of variable-stiffness shell structures
Student: Simon White
Supervisors: Paul Weaver and Giuliano Allegri (Imperial College London)
Composite materials enable the continuous variation of material stiffness properties in aero structures. The variable stiffness design philosophy has been proven to yield large performance improvements in flat structures. Its effect on the performance of shell structures however remains relatively unknown. For example, could such a design cancel the inherent buckling instability of certain shell structures? Key challenges in this area are: tackling the added complexity brought into the analysis by variable properties, and optimisation of components having large design spaces. This work will explore new and novel analysis techniques for the design of such structures; with particular emphasis on efficiency. The project will yield tools and insight for the design of refined shell structures
Simon was awarded a Faculty of Engineering Commendation for his thesis.