PhD and MSc by Research opportunities
Topics for PhD and MSc by research degrees in the Mechanics group.
Below are details of particular projects that could form the basis of PhD or MSc by Research in the Mechanics group. The appearence here of a project does not mean there is funding available (sadly) although if you are interested we would encourage you to contact us to discuss. Links to potential supervisors precede the decriptions of each project. The Head of the Mechanics group is Professor Jon Trevelyan.
Title: Modelling perivascular support using meshless methods
Supervisor: Dr Charles Augarde
Description: Modelling blood flow is of interest in the study of arteriosclerosis and diseases linked to heart problems. While many have modelled blood flow and the vessels themselves the surrounding (perivascular ) support to the vessels has been omitted (see here). The surrounding support may have as large an influence on flow as the properties of the vessels so omission could lead to major errors. The reason for its omission is the difficulty in modelling a semi-infinite support region which will be materially & geometrically nonlinear, and 3D. The approach we will take in this project will be to model this region using a meshless method which we have already demonstrated viable for 3D nonlinear problems in geomechanics and there is no reason why this cannot be applied to this problem. The project will involve investigation of suitable material models for the perivascular support, coding up of this into an implicit meshless code and a search for data against which to validate this approach
Title: Error estimation and adaptive strategies for the element –free Galerkin method
Supervisor: Dr Charles Augarde
Description: The element-free Galerkin method (EFGM) developed by Belytschko is perhaps the premier meshless method around today. To move meshless methods towards commercialisation in codes various problems need addressing (some of which we are currently working on here at Durham). Robust procedures for error estimation and consequent adaptive strategies are two of the items on the shopping list and these are the topics tackled in this project. We will begin by examining current error estimates in the EFG (there are not very many) and move on to development of a goal-oriented approach, as developed recently for the standard FEM and for XFEM. The project can be focussed on a number of areas, e.g. material and geometric nonlinearity, continua or fracture, but we will work in 3D from the start, making the project ambitious but exciting. There may be opportunities to work with leaders in associated fields such as Stephane Bordas at Cardiff University with whom we have links
back to start
Title: Parallel meshless modelling using GPUs with fast solvers
Supervisor: Dr Charles Augarde
Description: The full potential of meshless methods for solid mechanics will only be realised in 3D for very large problems (multi-million degrees of freedom). So how do we get there? If we are considering implicit approaches for static problems then serial approaches using direct solvers will become inefficient before we reach the level of detail we seek in 3D. Use of iterative solvers for serial processing is the first possibility to realise our goal in this project, working from previous publications and projects in collaboration with Ramage (Strathclyde). While this will provide extensive modelling possibilities on cheap hardware platforms a greater opportunity may exist with GPU-based machines, running in parallel. Here the challenges are (a) how to adapt the basic meshless method to run in parallel on GPUs and (2) what are good solution strategies in this environment? The project work in collaboration with colleagues in Computing Sciences in our joint School
Title: Developing the material point method for modelling tunnelling
Supervisor: Dr Charles Augarde
Description: The Material Point Method (MPM) is a mixed Eulerian/Lagrangian approach for modelling in solid mechanics bringing together ideas from finite element and particle-based methods. It was first developed in the 1990s as an explicit method for dynamic problems (a good summary is provided in Guilkey & Weiss, 2003, IJNME) but it can also be used as an implicit method for statics. Its advantage is its ability to model large deformations without the mesh entanglement issues met in finite elements while avoiding some of the current drawbacks of meshless methods (such as the EFGM). In this project we would look at using the method for 3D modelling of tunnelling (something I have been doing since my own PhD) looking at deformation prediction and assessment of stability. All the work will be in 3D from the start and the PhD could focus on the application or spend some time looking at improving aspect of the MPM itself
Title: Investigating the microstructures of compacted earth building materials
Supervisor: Dr Charles Augarde
Description: The use of earthen materials for modern structures is a hot-topic due to the inherent sustainability of these materials. If we could move these materials into the same position as held by concrete and brickwork, we might be able to make major reductions in the embodied energy of structures. This requires continued scientific research to understand behaviour. Most compacted earthen materials are mixtures of soil and water with the straw, lime or cement in certain cases. Understanding the porous nature of these materials is, we think, the key to understanding sources of strength, and the changes with humidity or temperature . We will look at these materials at the particle-level and develop measures of void size distribution based on fractal approaches and develop models of microstructure development with compaction or other processes (such as water evaporation from pores). These models may be built using stochastic procedures, acknowledging the heterogeneous nature of these materials
Title: Effective conservation of rammed earth via better understanding of material behaviour
Supervisor: Dr Charles Augarde
Description: Rammed earth is an ancient construction material in use for thousands of years. Many structures exist worldwide which are hundreds of years old. Their conservation is currently a concern to engineers and archaeologists and is heightened due to the effects of climate change. We recognise that effective conservation is only likely to occur if we improve our understanding of the engineering behaviour of this material, indeed many current conservation approaches have a negative effect due to lack of understanding, particularly of the role of water. We have worked with Chris Gerrard in Archaeology at Durham since 2004 on the development of effective strategies for conservation based on rigorous understanding of the material, which we see as a manufactured unsaturated soil, informed archaeology. This project will continue our work in this area and will focus on characterisation of heritage materials involving field work and laboratory testing allied to constitutive model development
Other projects in computational mechanics with Dr Charles Augarde
Supervisor: Dr Charles Augarde
Here are some more ideas ...
- Interpolations in the Material Point Method
- Development of fast fracture analysis using meshless methods and level sets on GPUs
- Solvers for meshless methods: what types of linear systems are produced and how to solve them?
- The effect of nodal arrangements & automatic grid generation for meshless methods
- Stress recovery procedures in meshless methods
- The scaled boundary finite element method for electrostatics/electromagnetics
- Adaptivity in meshless methods for 3D
Other projects in geotechnics and earthen construction supervised by Dr Charles Augarde
Supervisor: Dr Charles Augarde
Here are some more ideas, a mix of experimental and computational projects.
- Analytical and computational solutions for multiple tunnels in elastic ground.
- Geotechnical properties of Biochar materials
- Laboratory characterisation of Cob earthen construction materials
- Further lab characterisation of rammed earth using unsaturated soil mechanics
- Multigrid for geotechnical problems
Title: Strategic design co-ordination
Supervisor: Dr Graham Coates
Description:Design co-ordination has been recognised as a new and innovative approach to managing engineering design at an operational level. However, to be truly adaptive and responsive within a dynamic and changeable setting, the organisation and control of design must also be co-ordinated in a strategic sense. This research aims to develop appropriate strategic decision-making mechanisms related to the design domain, which are able to manage the potential iteration inherent in early design
Title: Engineering team composition
Supervisor: Dr Graham Coates
Description: Description: Engineering teams in companies consist of engineers from a range of disciplines with varying levels of capability in these disciplines. The aim of this research is to develop a means of determining and modelling engineers' skills and capability such that the composition of teams can be precision-planned against the project work to be undertaken. Further, the research will aim to deliver a method of identifying shortfalls in skills and capabilities within companies such that a development strategy can be formulated
Title: Integrated conceptual aircraft design
Supervisor: Dr Graham Coates
Description: Conceptual aircraft design is a highly complex process involving engineers from different disciplines and calculations of a multitude of inter-dependent parameters that define the design under consideration. The aim of this project is to research and develop a computer-based, integrated modelling framework to assist aeronautical engineers in the conceptual design stage of an aircraft. Within the context of a conceptual aircraft design environment, the framework developed will manage complexities such as change management and version control
Title: Nuclear reactor vessel structural integrity
Supervisor: Professor Roger Crouch
Description: Pre-stressed concrete pressure vessels provide the primary barrier for gas-cooled nuclear reactors. Their structural form can be complex; the many penetrations in the side wall and top-cap of the cylindrical vessel lead to local concentrations in stress and temperature. This project involves (i) extending an existing macroscopic constitutive model to include a thermal creep capability, and (ii) identifying a means of representing the many unbonded helical pre-stressing tendons within a 3D Finite Element code. With these new features, a series of coupled deformation analyses are to be undertaken to simulate the thermal-pressure cycling of a 40-year operational life. The project would suit an engineer with a strong interest in engineering mechanics and numerical analysis.
Title: Two-surface anisotropic hypoplasticity
Supervisor: Professor Roger Crouch
Description: A generalized anisotropic inelastic constitutive model for highly porous media with a collapsible fabric is yet to be developed. This project aims to construct such a model and develop a suitable stress integration algorithm to enable the deformation behaviour of collagen scaffolds to be simulated. The macroscopic hypoplastic formulation will be built from meso-level structural analyses which will inform the nonlinear hardening law. This work forms part of a larger study into the transport and deformation properties of biological tissue.
Title: Efficient implementation of the time-domain SBFE solution
Supervisor: Professor Roger Crouch
Description: The Scaled Boundary Finite Element approach offers an attractive alternative to the Boundary Element scheme when modelling the dynamic far-field in a Finite Element fluid-soil-structure interaction problem. Yet its use requires significant computational resource in a large 3D time-domain analysis (for example, if the convolution integral is used, nodal velocities for all previous time-steps are to be stored on the SBFE interface). This project aims to arrive at a more efficient implementation (making use of a parallel FE code) so that realistic structural analyses may be undertaken within reasonable runtimes.
Title: Parallel lattice Boltzmann model of temperature-driven multi-phase flow through deforming porous media
Supervisor: Professor Roger Crouch
Description: The investigation will build upon existing work in Durham on an LB formulation. The research will involve the exploration of different methods of simulating partially-saturated flow using a lattice approach. Here decisions will need to be made as to the manner in which the capillary surfaces are represented and The solid phase deformation handled. The final code will run on a large multi-processor machine to mimic the response of the transient thermal creep of CSH gel in the wall of a nuclear reactor vessel.
Title: Material instability in elasto-visco-plastic bodies
Supervisor: Professor Roger Crouch
Description: This theoretical study will explore the role of the intermediate principal stress on (i) the strength of porous media and (ii) the orientation at which rupture planes develop under multiaxial compression. The work will explore the relevance of the acoustic tensor (or more properly, the orientation where its determinant is minimized) as a measure of fabric changes under increasing deviatoric loading at differing levels of spherical compaction. The investigation will draw upon experimental work by others and earlier fracture (or failure) criteria to deliver a clear account of the transition from localised brittle rupture to distributed ductile flow as the degree of confinement increases and intermediate principal stress alters.
Title: Use of manganese oxides to treat contaminated land
Supervisor: Dr Karen Johnson
Description: Manganese oxides are powerful natural oxidants which are capable of immobilising many heavy metals and breaking down Persistent Organic Pollutants. This project involves working with mining companies whose ‘waste' tailings are in fact pure manganese oxides. The research would build upon ongoing research in this area and would investigate the use of these manganese oxides as contaminated land remediation products. The project would suit a soil scientist or geochemist with a strong interest in geochemistry and microbiology.
Title: Use of steel slag to treat contaminated water
Supervisor: Dr Karen Johnson
Description: Steel slag is a by-product of the steel industry and currently has limited markets. It has great potential for use in many environmental engineering applications including passive treatment of mine waters and landfill leachates which contain iron and manganese. This work would focus on the re-use of the spent substrates as either potential contaminated land remediation products or as further water treatment media for the breakdown of Persistent Organic Pollutants or the immobilisation of heavy metals. This research would build upon previous work in this area on the use of ‘wastes' to treat contaminated water and land. The project would suit an environmental engineer or geochemist with a strong interest in geochemistry and microbiology.
Title: Integrated Traffic Control for Freeway Networks
Supervisor: Dr Appie Kotsialos
Description: Traffic control measures are employed for the amelioration of traffic conditions and congestion elimination. Different control measures include ramp metering, route guidance and variable speed limits. The subject of this project is the design, development and implementation in simulation (macroscopic and microscopic) and in reality, of control strategies that are able to consider the concurrent operation of different kind of spatially distributed control measures avoiding any conflicting control actions. The development of a generic software tool able to address the integrated control problem for arbitrary-topology motorway networks is part of this study. This project is suitable for engineers with strong modelling, control and numerical optimisation background. Cooperation with motorway network site operators is expected for data collection and possible implementation.
Title: Modelling and Validation of the Motorway Traffic Flow Process
Supervisor: Dr Appie Kotsialos
Description: This project is concerned with developing a software tool that will be able to use different macroscopic traffic flow models (1st, 2nd or higher order) for modelling the vehicles' traffic in arbitrary topology motorway networks. Each of these models has a number of parameters that need to be estimated based on real data obtained from the network. Model validation should be used in each case employing different optimisation methods such as genetic algorithms, simulated annealing or other evolutionary method. The impact of operating different control measures, such as variable speed limits, should be taken into account. Furthermore, prediction algorithms for demand forecasting with origin-destination information should be incorporated. This project is suitable for engineers with strong modelling and numerical optimisation background. Cooperation with motorway network site operators is expected for data collection and possible implementation.
Title: Metal Spinning Process Simulation
Supervisor: Dr Hui Long
Description: Spinning is believed to have been first devised in ancient Egypt and since then has undergone transformative changes as a result of the advancement of modern CNC machines.Metal spinning has seen its increased applications for small batches of high quality components in medical, aerospace, defence, offshore and high value-added industries. One of the challenges in metal spinning is to prevent material failures during processing and to achieve required geometrical and dimensional accuracy, with sufficient production efficiency. By developing metal spinning process simulation techniques, we can explore the effects of process variables and material characteristics on the spun part quality so as to develop optimal spinning process strategies.
back to start box subsystem generally has a relatively low failure rate as compared to other subsystems. However, the gearbox has a low availability due to its high downtime per failure and gearbox failure incurs high costs for repair. To achieve a better efficiency of wind turbines and to reduce the cost of wind energy reliability prediction and failure mode analysis will be valuable for preventive maintenance of existing gearboxes. It is also essential for designing new gearbox to achieve reduced maintenance requirements and an increased reliability, especially for offshore wind turbines.
Title: Effective simulation techniques of dynamic springback in spinning
Supervisor: Dr Hui Long
Description: Springback is a phenomenon in which the plastically deformed metal sheet tends to return to its original position after unloading. In metal spinning, the sheet material deforms incrementally and repeatedly under the tool forces to produce the final geometry of the spun part. The springback not only occurs at the end of the spinning process but also takes place during the dynamic process of spinning. Simulation of the in-process dynamic springback requires high computational accuracy to modelling small dimensional changes of the part, while simulation of the material elastic-plastic deformation during spinning requires high computational efficiency to model the whole duration of the spinning process. The conflict of these two requirements makes existing Finite Element (FE) methods not feasible for the simulation of dynamic springback in spinning. This PhD study will investigate effective simulation techniques and integrate with existing FE implicit and explicit solution methods to modelling the dynamic springback in spinning.
Title: Effects of material cyclic loading on spinning
Supervisor: Dr Hui Long
Description: When a metal material undergoes plastic deformation under cyclic loading such as in spinning process, the material hardens until it responds more elastically in the original loading direction but softens when deformed in the opposite direction of the original loading. This phenomenon is called Bauschinger effect. The isotropic strain hardening model currently in use in our spinning simulation models could not count for this effect. Results show that there are unaccounted discrepancies between simulations and experiments. This PhD study will develop material constitutive models and FE simulation procedures to investigate the effects of cyclic loading of spinning on dimensional variations and product quality of spun parts.
Title: Simulation of crack failures in spinning
Supervisor: Dr Hui Long
Description: There are three main material failures observed in metal spinning process wrinkling, circumferential cracking and radial cracking. Wrinkling is a result of the development of high compressive circumferential stresses in the flange of spun metal which leads to buckling. On the other hand, high tensile radial stresses can lead to circumferential cracks whereas radial cracks can be a result from excessive circumferential and bending stresses FE and experimental based investigation of wrinkling is currently in progress. This PhD study aims to develop appropriate material damage models for FE spinning simulation models in order to investigate how the key process parameters affect the occurrence of cracks during spinning and to develop forming limit diagrams for preventing such failure.
Title: Reliability Prediction and Failure Mode Analysis of Wind Turbine Gearboxes
Supervisor: Dr Hui Long
Description: Gearbox is an important and expensive subsystem in an indirect drive wind turbine. The field reliability study of modern wind turbines shows that the gear
Title: Gear failure modes and reliability prediction
Supervisor: Dr Hui Long
Description: The design of conventional gear transmissions has been well documented in the International Standards. However, for a gear system used in wind turbine applications,the demand for gearboxes designed with extremely high gear ratio and operating under conditions subject to a broad spectrum of load and speed variations makes the reliability prediction and failure prevention difficult. In the absence of accurate field failure data of the different components in the gearbox, this PhD study aims to predict gear failure modes and reliability through probability analysis of key gear rating capacity factors. The stress levels and failure probabilities of pitting, bending, scuffing, micropitting and wear of gear teeth will be investigated. Spectrums of load and speed variations will be considered to determine duty cycle loading conditions for the probability analysis.
Title: Design tool for wind turbine gearboxes
Supervisor: Dr Peter Matthews
Description: The main functions of the gearbox are to transmit power from the low speed turbine shaft at high torque to the high speed generator shaft at low torque. Due to the required large transmission ratio, a gearbox typically consists of two or three stages, either planetary or parallel, to increase the speed. Various configurations of gear stages have been designed for wind turbine applications which show an improved reliability. The development of a gearbox design tool for wind turbine applications would be highly desirable to enable detailed design calculations and reliability prediction. Various gearbox design concepts using different configurations of planetary and parallel stages can be explored. This PhD study will introduce new methods in reliability prediction and develop new gear configurations based on an existing gear design software tool for wind turbine gearbox applications.
Title: Achieving Robust Design through stochastic support
Supervisor: Dr Peter Matthews
Description: Robust design is an area of increasing interest. Manufacturers benefit from being able to design and produce artefacts that will have a greater likelihood of being manufactured succesfully, as the artefacts have greater tolerance designed into them. This project will consider how to develop stochastic-based design methodologies to support designers into creating more robust designs, through the use of Bayesian modelling of the design domain. This project would suit an engineering graduate with strong interests in Artificial Intelligence methods or alternatively a statistics or computer science graduate with a general interest in design engineering.
Title: Agile Design
Supervisor: Dr Peter Matthews
Description: The concept of agility within the manufacture context is well known: it is the ability to rapidly respond to unexpected changes in the external environment. This project is part of an ongoing research theme to bring these ideas of agility to the design process. This project will investigate the nature of unexpected events within the design process and how these affect the design process. The aim is to deliver a means for creating a design environment that can respond to such unexpected events with a similar agility as in the manufacturing context. This project would suit an engineering or computer science graduate with an interest human factors and project management.
Title: Achieving Robust Design through stochastic support
Supervisor: Dr Peter Matthews
Description: Environmental legislation is resulting in modern production requirements stipulating for increasing documentation trails about the material and other resources used in manufacturing goods. The ISO-1404x standards provide guidance on the type and structure of the data that should be captured, but this is proving to be a challenge. This project aims to address these challenges by developing a resource based product model that can 'follow' material through its full life cycle. When coupled with Life Cycle Analysis tools, this will provide a powerful means for assessing the total impact a product will have on its environment. This project would suit either an engineering graduate or a computer science graduate with interests in knowledge management processes.
Title: Climate change impacts on geotechnical structures
Supervisor: Dr David Toll
Our climate is changing and we can expect more extreme rainfall patterns in winter months and longer, drier periods in summer months. Changes in the climate patterns will influence the seasonal variations in pore water pressures in the ground and this will have implications for serviceability and strength of geotechnical structures (road/rail embankments, dams, foundations for buildings etc). Large ranges of pore water pressure change between summer and winter will induce greater shrinkage and swelling in clay soils. There is also the possibility of increased slope failures (landslides) due to more intense storms. Possible PhD studies could involve:
(i) Field studies of pore water pressure and suction changes, involving a fully instrumented embankment
(ii) Numerical modelling of pore water pressure responses to climate change and the resulting volume change of the soil
(iii) Laboratory testing of soils using state-of-the-art suction controlled equipment for unsaturated soil testing
Title: XML data standards for Geo-engineering
Supervisor: Dr David Toll
The project will investigate ways to represent geotechnical data on the World Wide Web using XML. It will contribute to the GeotechML initiative (http://www.geotechml.com/) to develop internationally agreed data standards for geo-engineering information. It is frequently the case that data is passed between organisations (e.g. from site investigation contractor to consultant or client) in paper form (as a report). A more efficient way to exchange data is to use electronic data files that can be directly output from geotechnical database software and can be read into other geotechnical software packages. We also need to make data available on the World Wide Web e.g. case study data. PhD studies could involve defining the data standards, providing tools for managing the data or displaying the data on the web (e.g. using Java, SVG). The intention would be to develop an inventory of case studies for geotechnical structures (slopes, retaining walls, foundations) that can form the basis of a case-based reasoning system for preliminary design.
Title: Rainfall-induced Landslides
Supervisor: Dr David Toll
Most landslides are triggered by rainfall infiltration. As water infiltrates, pore pressures rise (or soil suctions decrease) leading to a loss of strength, leading to slope instability. In many tropical or arid regions, soils are initially unsaturated, so the soil behaviour is dominated by significant soil suctions. The process of infiltration is governed by complex unsaturated flow processes, requiring an understanding of the water retention properties of the soils, the relationship between permeability and suction (or degree of saturation). The relationships between strength, suction and degree of saturation are also complex. A project in this area could be experimental in nature, involving determinations of water retention behaviour, permeability functions and strength making use of the state-of-the-art unsaturated soil testing equipment available at Durham University. There is also scope for a project investigating the problem computationally, using finite element modelling of the unsaturated flow processes and hydro-mechanical behaviour.
Title: Insitu measurement of water content
Supervisor: Dr David Toll
In situ measurement of water content is an essential part of understanding soil behaviour in the field. While there are a number of devices available, they are bulky and the results can be affected by soil disturbance due to installing the device. This project will aim to develop a new device for water content measurement using the Time Domain Reflectometry (TDR) technique. The intention would be to investigate a coiled arrangement of electrodes that can be built into a smaller device than the conventional TDR probe, so that it can be used in conjunction with the Durham University high capacity tensiometer which is used for suction measurements. The new device would be calibrated by testing the device in the laboratory using a range of soil types (clays through to sands) over a wide range of water contents. The results would also be compared with on-going research in electrical resistivity methods for water content mapping.
Title: Engineering Behaviour of Unsaturated Structured Soils
Supervisor: Dr David Toll
Many natural soils are “structured” by the presence of chemical bonds and compacted soils are “structured” by the presence of a distinct fabric (arrangements of particles). Current constitutive models of soil behaviour are very limited in their ability to combine the effects of structure with unsaturated aspects. Nevertheless, many soils exist in an unsaturated state, particularly in tropical and arid regions of the world, but also in temperate regions. In an unsaturated state the behaviour is dominated by suctions and the behaviour becomes significantly more complex. PhD studies could involve laboratory testing of soils using state-of-the-art suction controlled equipment for imposing cycles of wetting and drying and also studying structure of unsaturated soils using Environmental Scanning Electron Microscopy.
Title: Interactive stress analysis of solids
Supervisor: Professor Jon Trevelyan
We have developed a software tool for rapid and interactive stress analysis in 2D (see http://www.conceptanalyst.com). The tool is based on the boundary element method (BEM) for reasons of speed and robustness of meshing and re-meshing. This has achieved a high level of interactivity because the speed of computation is sufficiently high that, when the geometry of a model is changed, the stress contours on the modified geometry can be updated in real-time. We would like to explore the potential for extension of these ideas to the analysis and re-analysis of 3D solids. We will investigate the extension of our fast numerical integration ideas from 2D to 3D, and determine substructuring, meshing and re-meshing strategies to optimise run-time. The project can be expected to make use of multi-core processing.
Title: Evolutionary structural topology optimisation of solids
Supervisor: Professor Jon Trevelyan
We have developed a set of algorithms by which a structure under a prescribed set of loads and displacement constraints may be optimised using the popular Evolutionary Structural Optimisation (ESO) method. In a recent PhD project, we have implemented this in both 2D and 3D using a BEM model superimposed on a B-spline boundary representation of the geometry (see Cervera & Trevelyan (2005), Computers & Structures, 83(23-24), 1902-1916 (Part I (2D)) and 1917-1929 (Part II (3D)). However, the 3D algorithm was only preliminary work to demonstrate the algorithm. This project will advance our understanding and development of a general algorithm for solids. It will require careful management of the boundary surface model and its evolution through the optimisation, and the development of appropriate strategies to optimise the convergence rate towards an optimum shape/topology.
Title: Interactive design analysis tool for crankshafts
Supervisor: Professor Jon Trevelyan
The crankshaft is among the most analysed of engine components. Its geometry is awkward for finite element meshing, containing toroidal fillets, oil drillings and various lightening holes and scalloped surfaces. The loading is also complicated, being determined by the cyclical motion, inertia forces and influenced by oil film thickness in the bearings. In this project we will develop an environment for the stress analysis of a crankshaft using a tailored boundary element pre-processor, and using the BEM re-analysis ideas we have developed in 2D simulations (see http://www.conceptanalyst.com).
Title: Integration schemes for 3D Partition of Unity BEM (PUBEM) in wave propagation simulation
Supervisor: Professor Jon Trevelyan
Our work in computational simulation of waves has become internationally recognised. In particular we focus on scattering problems in which the wavelength is considerably smaller than the size of the scatterer. The Partition of Unity is one of a number of approaches that have recently emerged that aim to overcome limitations of conventional FEM/BEM analysis in this situation. We have shown that the PUBEM is extraordinarily efficient in terms of the number of degrees of freedom required to solve such a problem, but is limited by the requirement to evaluate some oscillatory integrals. See http://www.dur.ac.uk/jon.trevelyan/ShortWavePage.htm. We have been exploring new strategies for computing these integrals and this PhD program will take that work further by considering a wide range of integrals that apply in different types of wave (acoustic, elastodynamic, etc.). This project would suit either an engineering or mathematics graduate.
Title: Potential for use of iterative solvers in Partition of Unity BEM (PUBEM) in wave propagation simulation
Supervisor: Professor Jon Trevelyan
Our work in computational simulation of waves has become internationally recognised. In particular we focus on scattering problems in which the wavelength is considerably smaller than the size of the scatterer. The Partition of Unity is one of a number of approaches that have recently emerged that aim to overcome limitations of conventional FEM/BEM analysis in this situation. See http://www.dur.ac.uk/jon.trevelyan/ShortWavePage.htm. We have shown that the PUBEM is extraordinarily efficient in terms of the number of degrees of freedom required to solve such a problem, but the governing matrix is ill-conditioned requiring a numerically expensive solver. However, recent work on circular scatterers in 2D has shown that if the model is defined appropriately then the conditioning improves to such an extent that much more efficient solvers can be used. Successful results have been obtained using Gauss elimination. For larger problems we will ultimately want to move to iterative solvers such as GMRES and Bi-CGSTAB. In this project we will explore the tuning of different PUBEM models to optimise their conditioning and move forward to investigate iterative solution. The work will involve the development of a preconditioning strategy for the particular character of matrix that is encountered in this method. This project would suit either an engineering or mathematics graduate.
Title: Comparison of different Partition of Unity BEM (PUBEM) formulations for short wave propagation simulation
Supervisor: Professor Jon Trevelyan
Our work in computational simulation of waves has become internationally recognised. In particular we focus on scattering problems in which the wavelength is considerably smaller than the size of the scatterer. The Partition of Unity is one of a number of approaches that have recently emerged that aim to overcome limitations of conventional FEM/BEM analysis in this situation. See http://www.dur.ac.uk/jon.trevelyan/ShortWavePage.htm. There are different ways that a boundary element approach may be formulated, though. One choice is whether to use the simpler and faster "collocation" method or the more stable "Galerkin" approach. Another choice relates to how we overcome the problem of non-uniqueness of integral equations at some frequencies - we could use the "CHIEF" method or the Combined Field Integral Equation (or Burton & Miller) approach. I would like to supervise a PhD student to undertake a detailed comparison of the different formulations. This project would suit either an engineering or mathematics graduate.
Title: Meshless methods for short wave propagation simulation
Supervisor: Professor Jon Trevelyan
Our work in developing innovative computational methods for wave problems has resulted in a new family of finite elements and boundary elements that overcome the traditional limitation that these elements can span no more than about λ/5 (λ is the wavelength under consideration). We are interested in extending the "Partition of Unity" ideas now to a meshless method using the MLPG approach of Atluri & Zhu, and implementing a plane wave expansion for the solution. This has the potential to offer a very efficient solution for wave scattering problems in inhomogeneous media.
Title: New Partition of Unity finite elements and boundary elements for short wave propagation simulation
Supervisor: Professor Jon Trevelyan
Our work in developing innovative computational methods for wave problems has resulted in a new family of Partition of Unity Method (PUM) finite elements and boundary elements that overcome the traditional limitation that these elements can span no more than about λ/5 (λ is the wavelength under consideration). Thus the elements can be very large in comparison with the wavelength. See http://www.dur.ac.uk/jon.trevelyan/ShortWavePage/htm. Accuracy of the method has been shown to be very good, but analysis of these errors shows the highest errors to accumulate around the ends of elements. We propose to develop new elements with shape functions based on a higher order of inter-element continuity to overcome this and improve the efficiency further for high frequency wave scattering. This will include the use of Hermitian and trigonometric shape functions in association with the PUM plane wav
Title: The use of 3D laser scanning techniques for freeform surface measurement and data Processing
Supervisor: Dr Qing Wang
Description: Product design is a critical function of the production system. The quality of the product design is probably the most important factor in determining the commercial success and value of the product. Quite often the quality inspection will need to be performed at several places in production. A sound measurement system provides an essential foundation to the production and consequently enables the enterprises to deliver high quality products and services. Over the past ten years, the 3D measurement system has been developed rapidly with the primary requirement from the aerospace and automotive industries for measurement accuracy and efficiency. The School has a state of art measurement instrument called the laser scanner which provides automated, non-contact measurement capability for large-volume applications. The aim of this project is to use laser scanning techniques to capture the complicated geometric shape of the parts, such as the irregular free-form curves on a turbine blade and their fine details. A series of experiments will be performed to investigate the measurement uncertainty in relation to the scan methods and scan speeds. It is expected that the student will be able to process the scan data efficiently in order to generate solid models, identify the operational features of the laser scanner and be able to develop product design rules that are compatible with the measured process capability of large and complicated components. The project would suit an engineering or computer science graduate with a strong interest in engineering measurement, reverse engineering, computer modelling and analysis.
back to start
Title: Prediction of disassembled automotive parts using artificial intelligence
Supervisor: Dr Qing Wang
Description: The disassembly of manufactured products when combined with the repair and reuse of reclaimed components is seen as an attractive option for reducing life cycle costs and environmental impact. In addition, in order to design and plan disassembly environments (i.e. flow processing lines), it is essential that information concerning estimated job times for disassembly operations are available. Such activities and their associated job times are directly related to the state of wear of individual components and sub-assemblies. The principal aim of this project is to investigate the application of Artificial Intelligence techniques for estimating the disassembly and repair work content of automotive components. The disassembly process will be reviewed and the disassembly model will be built, trained and tested using available test data collected from technical literature and the model will be validated to evaluate some of the input parameters on the chosen output. The project would suit an engineering or computer science graduate with a strong interest in artificial intelligent techniques, resource planning and product end of life design and development.
Title: Design and analysis of a sustainable logistic network
Supervisor: Dr Qing Wang
Description: Traditionally, marketing, planning, manufacturing, sales and distribution organisations along the supply chain have operated independently. Quite often these organisations have their own objectives and these objectives are conflicting. Many manufacturing operations are designed to maximize throughput, reduce lead time and lower costs with little consideration for the impact on inventory levels and distribution capabilities, with the result that there is not a single, integrated plan for the organization. Clearly there is a need for a mechanism through better supply chain design by which these different functions can be integrated. The aim of this project is to review the supply network design methods of the past 40 years and identify the factors involved in operating the supply network, the outsourcing strategy used, and the balance between supply and demand that affects the inventory of changes to the replenishment strategy. It is expected that a prototype supply network model would be established which is representative of the true nature of resource flows (materials, information, production etc) in the supply chain, hence reducing the cost of the network. The project would suit an engineering or computer science graduate with a strong interest in sustainability, logistic network modelling and optimisation.
Title: Through life cost management
Supervisor: Dr Qing Wang
Description: Traditionally, the product design process has tended to concentrate on functional parameters such as dimensional geometry and structural performance evolutions, and cost has generally been neglected, even though it is widely accepted that early implementation of cost analysis models influences the design changes of the product and provides explanations of the relationships between cost and design parameters. However, there are many features of a product that can be studied using a life cycle cost analysis model. The objective of this project therefore is to develop a life cycle cost model to demonstrate the cost breakdown structures in order to identify the feasible alternatives for a cost effective design of parts/products. It is expected that the analysis will provide the mechanism for measuring the criteria or indicators that will demonstrate the cost performance or acceptability of a product to the marketplace. The project would suit an engineering or computer science graduate with a strong interest in cost engineering and product life cycle modelling.
Title: Sustainable design and manufacturing
Supervisor: Dr Qing Wang
Description: In many past situations, environmental effects were ignored during the design stage of new products and processes. The challenge of sustainable design is to alter conventional design and manufacturing procedures to incorporate environmental considerations systematically and effectively. This requires embedding corporate sustainability goals in the product design process that considers full life cycle issues such as product take back or extended producer responsibility. This project aims to develop novel assessment methods for examining the relationships between product design, production process, costs and effects associated with the “triple bottom line” of sustainability. The project would suit an engineering or business graduate with a strong interest in corporate strategy and product life cycle design and development.
Title: Technology management for underpinning innovation in enterprises
Supervisor: Dr Qing Wang
Description: In today’s highly competitive manufacturing environment, there is a need to encourage enterprises to exploit new technology-led innovation strategies to ensure longevity in the face of increased competition. Enterprise Resource Planning (ERP) systems are the predominant business management systems found in most enterprises. The proposed research is to identify the critical technological needs of the enterprise and establish a framework to support the range of tools used for the management of innovation during a product’s life cycle. The project would suit an engineering or computer science graduate with a strong interest in enterprise modelling and computer aided engineering.
Title: Performance optimisation of modelling manufacturing systems
Supervisor: Dr Qing Wang
Description: Manufacturing systems are, by nature, dynamic systems due to the underlying random effects. By applying simulation techniques, the features and attributes of such systems can be understood greatly. During the past decades, a number of optimisation methods have been developed for automated system design, considering only a single objective. However, from a system designer’s point of view, it is very desirable to obtain optimal solutions considering all the objectives. This research will use an evolutionary approach to develop a multi-objective genetic algorithm for solving the flexible process sequencing problems. The project would suit an engineering or computer science graduate with a strong interest in genetic algorithms, simulation and manufacturing systems.