Eden Oration 2017
The Eden Oration 2017 was given on 1 December by Dr Graham Pullan, Reader and Staff Fellow in Engineering.
I am a Fellow in Engineering, or “nuts and bolts” as the late John Collier used to call it. Engineering is the application of science, mathematics, and creativity, to the development of technology for society. Engineering is intimately connected with the design of products, and systems, that all of us use in our daily lives. How can an Engineering academic, let alone one in this “Poor Society”, contribute to this technology pipeline? What does it mean to be an Engineering academic?
Engineering is a broad discipline and I emphasise that what I have to say is based on my own narrow experience. My field is aerodynamics and energy. The design of the carefully shaped aerofoil blades in a jet engine has been the thrust of my research career. The aim is to improve the efficiency of these machines so as to reduce the fuel consumption of jet engines, or increase the power output of electricity generation turbines. As well as fluid mechanics and thermodynamics, my research is also connected to the field of computing, as the simulation of the flow through an engine is a vital analysis tool.
The University does not make jet engines. The key to making a meaningful contribution to these products is to collaborate with companies that do. If you are thinking, perhaps, that Dr Pullan partners with industry so that his genius can use these companies as a one-way conduit to the marketplace beyond, I stress that this is not the case. Nor does the University bring the science while industry brings the funding and then waits patiently for the future to emerge. The interaction is two-way, technical and continuous. It is based on openness, respect and the pre-eminence of ideas. What does each partner contribute?
Industry brings knowledge, often acquired at great cost, of the development of ideas into products, especially the integration of systems into a functional whole. Industry brings experience of design, with all its inherent compromises and trade-offs. Industry pays close attention to the current market and also has one eye on the crystal ball for future changes in direction. Perhaps the most important point is that engineers in industry are exposed to the challenges and problems that are at the heart of the machines that they sell. Often, these engineers have in mind a concept that might address these challenges, but the pressures of current projects do not always allow space and time for development of these ideas. These same pressures, however, mean that Industry can out-run the University when a problem becomes urgent. In fact, the devices produced by Industry can sometimes precede the detailed understanding of the underlying phenomena that govern their operation; it has been said that “thermodynamics owes more to the steam engine than the steam engine owes to thermodynamics.”
Academic Engineers typically have knowledge characterised by deep expertise in specific areas. These individuals collaborate with each other, such that their aggregate technical base can be broad, but they do not have the range of skills, nor the infrastructure, needed to develop a product, an integrated system, and take it to the market themselves. Also, central to the University’s ability to contribute to Engineering problems is the community of students and post-doctoral researchers that constitute a remarkable body of talented, motivated individuals. This intellectual eco-system is our key asset and we must nurture this resource, fostering in the University an open environment where our latest ideas and concepts are discussed, challenged and improved.
We see that Engineers in Industry and Academia have complementary knowledge, skills, approaches and resources. When a collaborative group is assembled – a team – what characterises a successful partnership and what does Industry and the University gain? In my experience, successful collaborations are long-term, involve trust and mutual respect, and have active and continuous technical support for the project from within the company. A key step is the identification of the problem: the translation of a technical challenge in the commercial world to a question that needs answering in the research world. Once the question is known, the approach to answering it is far easier to determine.
Would it be easier if the project management chart was instead defined by funding arriving on day 1 followed by a long bar marked “research” and then an action to send the report back on day 1500? What is the return on the investment of time and energy that this University-Industry collaboration requires? The partnership is beneficial to both parties in three ways. First, the impact of the research is enhanced by the focus on relevant problems. Second, the collaboration provides a route to fast-track innovative concepts from research lab to product prototype. Third, the diversity of the background and skills of the participants provides fertile ground for the generation of ideas. My experience is that these collaborations have impact both in terms of success in academic publications and also in the generation of intellectual property.
I will now give an example of the arc of one particular idea that has been nurtured by the type of interactions I have described. The aerodynamic design of a jet engine is dominated by computational methods, many of them pioneered by our former Vice-Master, and now Emeritus Fellow, John Denton. As well as accuracy, Engineers in Industry are constantly craving reductions in computational run time. Reduced run time allows either more candidate designs to be evaluated, or more complex, and hence accurate, simulations to be performed in the same run time as existing software. The capabilities of computer hardware increase year on year, and this represents something of a free lunch for the developers of engineering software. But about 10-15 years ago, an idea for a step change in compute capability emerged. I say “emerged” because I cannot now remember the chronology, only a conversation with my late brother, a professional software developer, confirming that the following was worth a shot.
Computers contain two types of processor. One is general purpose and handles everything from the operating system to the word processing software with which I typed this speech. The other processor is specialised and its function is only to provide rapid display of complex graphics. 15 years ago, the demand for physical realism, from players of computer games, was such that software developers had begun co-opting the graphics hardware, designed to rapidly colour in triangles on the screen, to perform rudimentary fluid dynamics effects such as smoke and water flow. The challenge was whether sophisticated, engineering-quality, simulation software could also be made to run on these chips. If so, aerodynamicists would have access to the far more powerful graphics processor (current graphics chips have 4000 cores, compared to the 18 of the most advanced general purpose processors).
This challenge was enthusiastically met by my PhD student, Tobias Brandvik (also a member of this College). I can skip 4 years of research by summarising that the new software was 10 to 20 times faster running on graphics processors than the original code running on general purpose chips. This immediately opened up wide-ranging opportunities for turbine aerofoil design. Industry, having followed our progress with interest, was quick to see the possibilities. A variety of path-finding projects were initiated on design optimisation, interactive virtual reality design systems and large scale simulation. But there was also strong encouragement from Industry to tackle one of the grand challenges in the field: the prediction of compressor stall.
Just as aeroplane wings lose lift at high angles of attack, so compressor aerofoils can stall, losing pressure rise and thrust capabilities. This is a serious event for a jet engine, leading to shut down of the machine. Fellows and Scholars will be relieved to learn that the operation of the compressor is therefore limited, by a “stall margin”, such that stall does not occur in flight. Although characteristic signatures of stall inception had been recorded in many experiments, the fluid mechanic origins of stall remained illusive. In a three-way collaboration between Cambridge, Mitsubishi Heavy Industries and MIT, we used our enhanced computational capability as a virtual experiment in which all parts of the compressor flow could be interrogated. By asking the right questions, we were able to determine not only the root cause of the phenomenon but also to develop new compressor designs with extended, un-stalled, operating range. This successful translation of new capability, to fluid mechanic learning, to engineering product, would not have been possibly without University and Industry working closely together.
In the lifespan of Trinity Hall, Engineering is a young discipline but one that has had an enormous impact on all our lives. I like to think that Dr Eden’s provision of candles for the Chapel suggests an interest in combustion and heat transfer, even though he wrote his Will 200 years before Joule’s experiments lead to the determination of the First Law of Thermodynamics. I have said that Engineering is the application of science, maths and creativity to the development of technology and I would like to end by emphasising the value of the working environment in the stimulation of innovation. For centuries, Trinity Hall has provided an environment for the exchange of views, for the generation of ideas, and for the nurturing of these fragile concepts into the firm foundations that others may build on. I am grateful to continue to be inspired by, and learn from, all of those that form this Poor Society, as well as my friends and colleagues in other academic institutions and in Industry. At least in my career, the whole has been greater than the sum of the parts.
Before I close, it is customary to reflect on recent changes to the Fellowship. A number of Fellows have left Trinity Hall during the past year: Dr Libby Caygill, Professor Jane Clarke, Dr Amaleena Damlé, Dr Felix Deschler, Ms Di Haigh, Dr Lindley Lentati, Dr Gunnar Möller, Dr Poornima Paidipaty, Dr Paul van Pelt, Dr Tadashi Tokieda and Dr Jens Zimmermann. It was with much sadness that we learned of the death of David Fleming, Emeritus Fellow, in March this year. David’s contribution to Trinity Hall, over 40 years as a Fellow in Law and Tutor, was remarkable. Many of us will have fond memories of our interactions with David, my own are of conversations on cricket where his sharp legal mind, and powers of recall, were applied to the fortunes of the England team.
I am delighted to be able to report the arrival of several new members of the Fellowship: Dr Gonçalo Bernardes (Fellow in Chemistry), Dr Koen Jochmans (Fellow in Economics), Dr Ron Reid-Edwards (Fellow in Mathematics) and Dr Daniel Tyler (Fellow in English); Research Fellows – Dr Guillermo Burgos Barragan, Dr Eugenio Giannelli and Dr Nicola Kozicharow; Fellow Commoners – Mr Jai Chitnavis, Dr Aled Davies and Professor James Ritter. And The Rt Hon Sir David Bean and Mr Andrew Marr have become Honorary Fellows.
Dr Eden’s kind provision of “Wine and Diet” now awaits us. I hope that you will enjoy the evening and, in particular, your conversations with the other guests. You never know what possibilities for inspiration, or perhaps even future collaboration, await.