Term has started, and our students are preparing for end-of-semester examinations; so, I suspect that they would welcome the opportunity to deploy the sleeping-learning that Aldous Huxley envisaged in his ‘Brave New World’ of 2540. In the brave new world of digital engineering, some engineers are attempting to conceive of a world in which experiments have become obsolete because we can rely on computational modelling to simulate engineering systems. This ambitious goal is a driver for the MOTIVATE project [see my post entitled ‘Getting smarter‘ on June 21st, 2017]; an EU-project that kicked-off about six months ago and was the subject of a brainstorming session in the Red Deer in Sheffield last September [see my post entitled ‘Anything other than lager, stout or porter!‘ on September 6th, 2017. The project has its own website now at www.engineeringvalidation.org
A world without experiments is almost unimaginable for engineers whose education and training is deeply rooted in empiricism, which is the philosophical approach that requires assumptions, models and theories to be tested against observations from the real-world before they can be accepted. In the MOTIVATE project, we are thinking about ways in which fewer experiments can provide more and better measured data for the validation of computational models of engineering systems. In December, under the auspices of the project, experts from academia, industry and national labs from across Europe met near Bristol and debated how to reshape the traditional flow-chart used in the validation of engineering models, which places equal weight on experiments and computational models [see ASME V&V 10-2006 Figure 2]. In a smaller follow-up meeting in Zurich, just before Christmas [see my post ‘A reflection of existentialism‘ on December 20th, 2017], we blended the ideas from the Bristol session into a new flow-chart that could lead to the validation of some engineering systems without conducting experiments in parallel. This is not perhaps as radical as it sounds because this happens already for some evolutionary designs, especially if they are not safety-critical. Nevertheless, if we are to achieve the paradigm shift towards the new digital world, then we will have to convince the wider engineering community about our novel approach through demonstrations of its successful application, which sounds like empiricism again! More on that in future updates.
Image by Erwin Hack: Coffee and pastries awaiting technical experts debating behind the closed door.
Six months ago I wrote about our EU research project, called INSTRUCTIVE, and the likely consequences of Brexit for research [see my post: ‘Instructive report and Brexit‘ on March 29th, 2017]. We seem to be no closer to knowing the repercussions of Brexit on research in the UK and EU – a quarter of EU funding allocated to universities goes to UK universities so the potential impacts will hit both the UK and EU. Some researchers take every opportunity to highlight these risks and the economic benefits of EU research; for instance the previous EU research programme, Framework Programme 7, is estimated to have created 900,000 jobs in Europe and increased GDP by about 1% in perpetuity. However, most researchers are quietly getting on with their research and hoping that our political leaders will eventually arrive at a solution that safeguards our prosperity and security. Our INSTRUCTIVE team is no exception to this approach. We are about half-way through our project and delivered our first public presentation of our work at the International Conference on Advances in Experimental Mechanics last month. We described how we are able to identify cracks in metallic structures before they are long enough to be visible to the naked eye, or any other inspection technique commonly used for aircraft structures. We identify the cracks using an infra-red camera by detecting the energy released during the formation and accumulation of dislocations in the atomic structure that coalesce into voids and eventually into cracks [see my post entitled ‘Alan Arnold Griffith‘ on April 26th, 2017 for more on energy release during crack formation]. We can identify cracks at sub-millimetre lengths and then track them as they propagate through a structure. At the moment, we are quantifying our ability to detect cracks forming underneath the heads of fasteners [see picture] and other features in real aerospace structures; so that we can move our technology out of the laboratory and into an industrial environment. We have a big chunk of airplane sitting in the laboratory that we will use for future tests – more on that in later blog posts!