Numerical shape optimisation with CFD is an area of strong industrial interest, as demonstrated by the successful EU projects FlowHead and About Flow: given 'a set of shape modes (parametrisation) that the can be modified, the algorithm finds the best combination of modes to achieve a minimal cost function such as drag, or pressure drop. The choice of parametrisation, i.e. the number and type of degrees of freedom that the algorithm can vary, is of course crucial for the optimised performance. Good optima will typically require very rich design spaces which are complex and labour-intensive to set up. Even when the optimal shape is found, crucial features may be lost when translating the shape back into a CAD-format that can be further processed with other tools.

There is a range of possible parametrisations, of particular interest for industrial application is the CAD-based approach developed at QM, where the optimal shape is a) automatically derived from an existing CAD setup without additional setup, and b) is presented in the end as a surface in CAD format again, which can either be manufactured directly, or be further analysed in the virtual prototyping chain. This approach is currently the only one in the literature that works with the CAD description inside the design loop, and is robust enough for industrial application. A paper has been published with a proof of concept using an implementation that has a limited vocabulary of CAD shapes.

The proposed research project develops this further to 
1. increase the vocabulary of the implemented CAD kernel to include trimmed NURBS patches, fillet radii, etc.
2. incorporate of manufacturing constraints such as minimum thickness, box constraints, minimum radii, casting deforming constraints
3. adaptively refine the NURBS patches of the CAD shape where the design is sensitive and more control is required
4. integrate with the alternative non-CAD parametrisation based on surface grids to achieve a seamless environment capable of tackling a wide range of applications.
5. demonstrate the framework on a variety of fluid-mechanic, structural and multi-disciplinary problems.

The resulting software package will be of interest to a range of industrial applications, from turbo-machinery as initially envisaged by the co-sponsor, to automotive, Formula 1 or aeronautics.

Supervisors: Dr. J.D. Mueller, Prof. V. Toropov

QMUL Research Studentship Details
• Available to Home/EU/International Applicants
• Full Time programme only
• Applicants required to start in September/October 2015.
• The studentship arrangement will cover tuition fees and provide an annual stipend (£15,863 in 2014/15) for up to three years.
• The minimum requirement for this studentship opportunity is a good Honours degree (minimum 2(i) honours or equivalent) or MSc/MRes in a relevant discipline.
• International applicants should refer to the following website at http://www.qmul.ac.uk/international/index.html 
• If English is not your first language then you will require a valid English certificate equivalent to IELTS 6.5+ overall with a minimum score of 5.5 in all sections (Reading, Listening, Writing, Speaking).

Contact Details:
For informal enquiries about this position, please contact Dr. J-D Mueller 
Tel: 020 7882 5421
E-mail: j.mueller<στο>qmul.ac.uk

Application Method:
To apply for this studentship and for entry on to the PhD Mechanical Engineering - Semester 1 start (between Sept - Dec) via the following webpage: 
http://www.qmul.ac.uk/postgraduate/pgrcoursefinder/engineering/index.html

Further Guidance: 
http://www.qmul.ac.uk/postgraduate/applyresearchdegrees/index.html

Please be sure to include a reference to ‘2015 QMRS SEMS JDM’ to associate your application with this studentship opportunity.

Deadline for Applications: 27/02/2015