Extended projects 2015 cohort

3D printed patient-specific revision hip implants

Behjat Ansari Student: Behjat Ansari
Supervisors: Kate Robson Brown, Mark Schenk and Richard Trask (University of Bath)

Current hip implants are not tailored to each patient, thereby causing unnecessary removal of healthy cancellous bone in and around the acetabulum to accommodate them, prompting the need for revision surgery to restore the natural geometry. This study evaluates the need for creating bespoke revision hip implants, through the characterisation of the trabecular architecture of the acetabulum, finite element analysis of a human innominate bone under typical load cases, and identification of suitable substitute material, together with processing methods.

The need for bespoke implants is evident throughout the study and the literature, from the variations in the trabecular architecture of different hip specimens to variations in stress distributions of different finite element models. Thus, a bespoke solution is proposed to address these differences and consequent hip implant failures which is now made possible with advancements in material and processing technologies, most notably additive manufacturing and the ability to produce porous scaffolds.

Back to top

Development of improved fibre reinforced feedstocks for high performance 3D printing

Lourens BlokStudent: Lourens Blok
Supervisors: Kevin Potter, Ben Woods, Marco Longana, HaNa Yu and Amir Rezai (BAE Systems)

Composite materials made from carbon fibres and polymer matrices can provide excellent mechanical properties and allow for significant design tailorability. A fundamental challenge with fibre reinforced plastics, however, is the combination of the reinforcements into the polymer matrix with good consolidation, maximum control of fibre orientation and low cost. In this work, an additive manufacture (AM) approach was used where the fibres are first embedded in a thermoplastic matrix in a pre-processing stage. The novel aspect of the project is that the continuous fibres were replaced by highly aligned short fibres to allow for improved printing performance and increased design freedom. Careful selection of the fibre length to be longer than the critical length of the fibre/matrix system allows for retention of the majority of the mechanical performance of a continuous fibre solution, without the need for cumbersome and restrictive fibre cutting and initial laydown procedures within the additive manufacturing process.

Two main advantages are envisioned to follow from this project: (i) the process will enable 3D printing of (short) fibre composites parts with the highest reported mechanical properties to date and (ii) the process is better suitable for recycling by using short fibres and a thermoplastic resin. These advantages may lead to a broader range of applications for 3D printed parts, where higher mechanical properties are required whilst keeping the advantage of rapid prototyping and customisability.

Steps in the development of short fibre reinforced filament

Back to top


Surface modification of composite material for space applications - a baseline

Desmond HeStudent: Yanjun He
Supervisors: Ian Hamerton, Mark Schenk and Alex Brinkmeyer (Oxford Space Systems)

One of the challenges of using flexible deployable structures is to demonstrate their ability to resist the extremely harsh environment of space – the combination of deep vacuum, UV, proton and electron radiation and atomic oxygen interaction can strongly affect the mechanical performance of the material. In addition, the extreme temperature fluctuations encountered in space can lead to thermal distortion, degradation and/or failure.

One of the possible solutions for reducing the degradation of flexible composites structures due to the space environment is to apply surface modification techniques and/or coatings. Coatings and surface modification such as the Photosil™ process have been used on composite materials to counter the harsh space environment. However, in some space applications such as deployable structures, tests should be performed to ensure the modifications or coatings do not affect the properties of the material.

This project aimed to create a baseline for the composite materials, in preparation for environmental exposure tests. Four kinds of composite materials were investigated. The investigation was conducted by three main tests, which were three-point bending, bend radius test and dynamic mechanical analysis (DMA). Results provide a baseline for the future testing of the materials to evaluate their suitability as space deployable structures.

Back to top

Accuracy and efficiency of the finite element method at modelling discontinuities in composite laminates

Aewis HiiStudent: Aewis Hii
Supervisors: Luiz Kawashita, Alberto Pirrera, Stephen Hallett and Adam Bishop (Rolls-Royce)

The finite element method (FEM) is increasingly being used for design and virtual testing of composite laminates, where accurate and efficient simulation of delamination phenomena is required. Delamination introduces a strong discontinuity in the numerical domain, where the FEM solution loses accuracy. Commercial finite element codes offer a plethora of element formulations for generic application, however there is no informed guidance on applicability of these elements for delamination analysis in composite structures. With the limited computational power and element formulations in commercial FE codes, it is imperative for structural analysts to achieve good trade-off between accuracy and efficiency in modelling delamination.

The aim of the six-month project was to study the accuracy and efficiency of different element formulations at modelling both static and dynamic behaviour of composites with an embedded discontinuity. The investigated element types were ‘continuum shells’ and hexahedral continuum elements with various order of interpolation, integration and augmentation functions. Numerical experiments were carried out on plates subjected to mode I, II and mixed mode I/II loadings. The accuracy of each element type was measured by the error in the strain energy release rate (computed via the Virtual Crack Closure Technique (VCCT) and the J-integral calculation) as compared with 20 node hexahedral element with full integration. The efficiency of the elements was measured by the CPU time required by each element type.

The study found consistently that elements that can develop curvature upon bending are naturally more accurate near the crack tip. Bending curvature can be achieved without high penalty in computational costs, via shape function augmentation on the ‘cheaper’ linear elements or utilising kinematics of shell elements. On the other hand, the dynamic test cases demonstrated that elements which are inaccurate near the discontinuity will accumulate errors as the solution is time-marched. Consequently, the final state of the model can have very different dynamics (i.e. magnitude and phase of the response) compared with the more accurate model.

Back to top

Structural efficiency of a beam model through stiffness adaptation

Olivia LeaoStudent: Olivia Leão Carvalho
Supervisors: Alberto Pirrera, Rainer Groh and Paul Weaver

It is common practice to design stiff structures that exhibit small deformations as a way of avoiding nonlinearities and simplifying the design process. However, this recurring design philosophy is inconsistent with the challenges faced by modern engineering, often driven by structural efficiency. As computational power improves and nonlinear analyses become cheaper and more advanced, nonlinearities can not only be better understood but also exploited for a benefit.

This research proposes looking into nonlinearities from a different perspective in which they are embraced and their capabilities explored in the design process. This approach is anticipated to deliver superior efficiency for various structural members under the claim that restricting structures to behave linearly may correspond to overdesigning them in cases where deformation constraints are relaxed. A simple T-beam with corrugated web is explored to demonstrate the concept of efficiency via stiffness adaptation. The proposed structure relies on a hidden length mechanism to induce the desired nonlinear response and the effect of several geometry parameters are investigated through a sensitivity analysis study.

Back to top

Design of small scale bend-twist coupled wind-turbine blade demonstrator

Vincent MaesStudent: Vincent Maes
Supervisors: Alberto Pirrera, Terence Macquart and Paul Weaver

The ever increasing energy demands and drive for eco-friendly technologies has fuelled an acceleration in research and development over the past decades. Within the wind energy sector, larger blades are being designed to extract more energy and improve overall cost efficiency. Larger blades, however, suffer significantly under the effects of gusts and general fatigue loads. These loads result in an effective limit in either life-time or maximum size and hence restrict the energy capacity of the blades. In order to allow for larger and more efficient blades, gust load alleviation has been identified as a much needed development. Bend-twist coupling, when designed to provide twist to feather, offers a passive solution. While significant research has been done on various aspects, penetration into market has not yet occurred. This can be in part attributed to only a small number of demonstrators being manufactured thus far, limiting the available test data for model validation.

The current work aims to assemble the basic components needed to design such demonstrators efficiently, such that they may be built and scaled up. Within this, use will be made of shell elements, in contrast to most literature where beam elements are used. The use of shell elements is expected to provide the models with an increase in fidelity crucial to accurately capturing the bend-twist coupling behaviour. The final aim of the project is to accelerate commercialisation and wide spread use of bend-twist coupled blades. In the process many questions, regarding best design practicing and manufacturability considerations, will be addressed.

Back to top

Novel patterned composites with functionalising additives by 3D printing

Arjun RadhakrishnanStudent: Arjun Radhakrishnan
Supervisors: Dmitry Ivanov, Ian Hamerton and Milo Shaffer (Imperial College London)

The project was aimed at determining the feasibility of improving the through-thickness conductivity of non-conducting continuous fibre composites using carbon nanotubes (CNTs). Conventional liquid composite moulding (LCM) processes are hindered by the high viscosity of CNT suspensions and filtration of nanoparticles through the fabric. Hence, two novel approaches were explored in conjunction with each other to overcome the limitations of conventional methods: a) use of heterogeneous solution developed in Imperial College London and b) liquid resin printing (LRP) which was developed here in the University of Bristol.

Figure 1
Modified Reprap printer – Liquid resin printer

Figure 2
Functional resin patch printed using LRP

Figure 3
Optical micrograph of heterogeneous solution

The viscosity of the heterogeneous suspensions were tailored to LRP and injected into preforms using optimised processing parameters, primarily the volume, rate of injection and delay between each injection. Further, the effect of the consolidation pressure and time of its application on the morphology of the resin patch was studied. This was conducted in order to characterise the filtration of the particles and the suppression of voids.

The study has shown that high CNT loadings of up to 3.5% can be achieved via a combination of these novel approaches. The through-thickness electrical conductivity was shown to exceed the previously reported values for non-conducting composites manufactured using LCM techniques. The project also demonstrated the feasibility of improving local properties.

Back to top

Unidirectional hybrid composite overload sensors - robust tools for visual overload indication

Tamas RevStudent: Tamas Rev
Supervisors: Michael Wisnom, Gergely Czel and Meisam Jalalvand

A novel, purpose-designed, thin interlayer glass/carbon hybrid composite sensor concept is presented that can be used for structural health monitoring (SHM) purposes in composite structures, leading to a safer operation in service. The UD hybrid sensors indicate the overload of a certain structure by exhibiting a change in appearance when loaded over a predefined strain value. The originally intact carbon layers absorb the incident light through the translucent glass layer showing a dark appearance as seen in Figure 1 (a). After the strain exceeds the failure strain of the carbon layers, the light is reflected back from the damaged glass/carbon interface exhibiting light stripes around the cracks in the carbon layer as illustrated in Figure 1 (b). These robust and lightweight sensors are completely wireless, do not require any data acquisition or evaluation system so they offer low-cost and simple solutions for visual strain overload indication. An analytical model developed here allows for tailoring the sensors to suit different substrate stiffnesses and strain margins. The sensors can be attached to a component either as a structural sensing layer or integrated locally to the structure as demonstrated through real-life applications. Furthermore, various test methods have been used to characterize these sensors including mechanical testing, DIC and AE measurements. An optimal configuration has been found while excluding the short type sensors as they exhibited a significant error in their trigger strain. The sensors can potentially be applied in advanced lightweight applications such as sporting goods, civil engineering structures (e.g. truss and bridge elements) as well as pressure vessels.

Figure 1 The sensing mechanism behind the UD hybrid composite strain overload sensors: (a) intact carbon layers absorbing light at glass/carbon interface (b) striped pattern visible due to light being reflected back from locally damaged glass/carbon interface around the sensing layer cracks

Back to top


A two-dimensional analytical model of the fish bone active camber concept using laminated composite plates

Andres RiveroStudent: Andres Rivero Bracho
Supervisors: Ben Woods, Paul Weaver and Jonathan Cooper

The Fishbone Active Camber (FishBAC) concept is a morphing trailing edge device that was developed for generating large, continuous and smooth changes in camber distribution, to gain lift control authority at a ‘low’ drag penalty. Initial structural and aerodynamic analysis of the FishBAC was performed by coupling a one-dimensional structural formulation (i.e. Euler-Bernoulli beam theory) with a two-dimensional aerodynamic solution (i.e. Vortex Panel Method) and the first prototype was manufactured by 3D-printing techniques using ABS polymer.

FishBAC morphing trailing edge concept

Despite having a low computational expense, a beam model does not allow to capture variations in out-of-plane displacement in two-dimensions, which are proper of three-dimensional aircraft fixed wings. Furthermore, a one-dimensional model limits the possibility of exploiting material and structural anisotropy for stiffness tailoring in both chordwise and spanwise directions. This project focuses on developing a two-dimensional structural model of the FishBAC using Kirchhoff-Love Plate Theory and Classical Laminate Theory that captures the static behaviour of a composite FishBAC morphing device and accounts for stiffness discontinuities due to the presence of stringers. This model will then be coupled with a three-dimensional aerodynamics solution and will be used to manufacture the first composite FishBAC morphing device.

Back to top

Development of light-weight, high mechanical strength carbon fibre composites suitable for turbine blade applications

Beth RussellStudent: Bethany Russell
Supervisors: Ian Hamerton, Carwyn Ward and Shinji Takeda (Hitachi Chemical)

Glass fibre reinforced composites are widely exploited in the manufacture of wind turbine blades. Epoxy resins in this application have been widely used due to their inherent high specific strength. A new infusible epoxy resin system has been developed, by Hitachi Chemical Co. Ltd, which offers enhanced toughness, lower weight, high strength and particularly good interphase region between the resin and fibre. In this work extensive studies on the epoxy-anhydride resin system were conducted, allowing the determination of thermal properties, such as the glass transition temperature and modulus, along with the cure reaction kinetics. Processing parameters of rheological behaviour and in-situ resin shrinkage during cure were also studied. The interlaminar shear strength and resistance to drop weight impact of a laminate was also assessed. The apparent interlaminar shear strength was determined to be 64.4 MPa, verifying the improved interlaminar properties as originally stated. The multifaceted nature of this research enabled the characterisation of a novel composite system which has been pivotal in identifying the areas of the system which can be developed further in the future.

Back to top

Industrial scale nano-reinforced composite structure: controlling delamination through vertically aligned carbon nanotubes

Rob WorboysStudent: Robert Worboys
Supervisors: Luiz Kawashita, Ian Hamerton, Stephen Hallett and Rob Backhouse (Rolls-Royce)

The aim of this study was to investigate the effectiveness of vertically aligned carbon nanotube interleave inserts (known commercially as Nanostitch™) as delamination crack suppressors within CFRP composite laminates. Nanostitch™ has been suggested by its supplier to enhance the interlaminar fracture toughness of a laminate, through bridging of the resin rich interlaminar region. However, to date, no independent research has been published, which investigates its behaviour in mode I or mode II fracture toughness tests. This study was a preliminary experimental and numerical investigation of non-interleaved baseline laminates. Repeatable double cantilever beam (Mode I) and cut-ply (Mode II) experimental tests have indicated that the baseline data collected is valid. Results are also consistent with verified finite element models Therefore, these baseline cases are ready for future experimental comparisons with similar interleaved laminates and numerical models.

Figure 1 The test set-up for double cantilever beam (left) and cut-ply (right) experiments in order to measure the mode I and mode II fracture toughness respectively.

Back to top

Edit this page