Please fill in the form below and we will get back to you shortly.
University of Huddersfield
Daresbury, Warrington, UK
28 February 2020
Project Reference – WALCE29
Main Supervisor: Prof. David Walker
Co-Supervisor: Dr Guoyu Yu, Dr Hongyu Li
Research into innovative methods of measuring complex surfaces, supporting autonomous manufacturing for science and industry.
The science of measurement is key to precision manufacturing in a global marketplace. Particularly challenging is the measurement of complex surfaces for lenses and mirror (e.g. for head-up and head-mounted displays, remote-sensing from space, etc), and for prosthetic joint-implants, turbine-blades, industrial moulds & dies, and much else besides. Methods for processing such surfaces are well established, the measurement is the limiting factor. This project will research and develop a novel measurement method, with the principal goal of delivering an integrated solution to be incorporated in an autonomous manufacturing cell – which will be a world first.
Technologies to machine and polish free-form optics are well developed. However, the weak link is the limitations in measuring these surfaces to “optical precision”, to close the process-loop during manufacture, and to certify final conformance to specification afterwards. The end-point, depending on the application, maybe a component from centimetres to metres in size, with surface-precisions from a few microns down to nanometres. The field is probably unique in finishing macroscopic components down to atomic dimensions.
This PhD project builds on two stages of funding from the EPSRC Future Metrology Hub and seeks to take the next major step forward in developing a novel free-form metrology instrument. The current status is that a prototype instrument with partial functionality has been built and demonstrated, operating under a much simplified mathematical model, and with instrument control within the Matlab software environment. The former limits precision, and the latter data-acquisition speed.
The successful PhD candidate will first develop a rigorous mathematical basis for the instrument, and then explore the underlying theory of calibration and uncertainty-of-measurement. This will lead into software development for real-time instrument control. Whilst ‘proof of concept’ coding in Matlab is a perfectly viable approach, this will need to be re-coded in a language such as C++ to achieve adequate dynamic performance and sampling speeds.
The methodology and software for instrument calibration and data-analysis will then be perfected using measurement-data acquired on real test-parts. Specific calibration-artefacts will be chosen which are easy to measure independently, but which can be specially configured to exhibit free-form characteristics. Such components will be measured in the lab by standard methods (e.g. interferometry), and results compared with the new instrument. This information will then be used to validate the uncertainty of measurement and, in particular, any systematic errors. Results will then be used to refine the mathematical and software basis of the instrument and methods for alignment, and thus optimize overall instrument performance.
There will be opportunities to examine other measurement techniques and compare their limitations with the new instrument above. The resulting thesis should be ‘nicely rounded’, with the bulk devoted to the mathematical basis and real-time software development, but supported by experimental results and their interpretation. Beyond acquiring skills in great demand, such as software, mathematics and data-acquisition, qualified metrologists are in short supply. Following a successful PhD, the student should, therefore, be assured of ample and rewarding employment opportunities.
The student will have a first-class or upper second class degree in a relevant engineering, mathematics or physics discipline. A good mathematical grounding is essential, particularly in areas relevant to establishing the complex geometric relationships in the instrument, and the statistical basis of the uncertainty of measurement.
Also, some competence in scientific programming would be very useful, which may usefully include real-time applications, C++ and Matlab, or the basic skills and enthusiasm to learn them. Experience programming recreational computer-games – another time-critical application – could well prove relevant.
If the student has experimental skills and is inclined towards performing independent measurements of test-artefacts, that will be encouraged, but this is not a requirement.
****This might be relevant to STFC staff through a secondment to a PhD programme, where the project-scope is “tuned” to be of mutual interest****