|
Interview with K Andrew Cliffe,
University of Nottingham, UK
by Hinke Osinga,
University of Bristol, UK |
Andrew Cliffe in his office, April 2010. |
|
Andrew Cliffe is Professor of Computational Applied
Mathematics at the University of Nottingham. He obtained a BA in
Mathematics at the University of Cambridge in 1975 and subsequently
got a position as a Scientific Officer at the Atomic Energy Research
Establishment, at Harwell Laboratory. Hinke Osinga asked what motivated him to
work in industry, and why after 30 years, he decided to apply for a
Professorship.
Working in a research lab in industry
As a student Andrew loved
mathematics. During his degree he chose to study topics on the applied
side with courses in Fluid Mechanics, Electromagnetism, Elasticity,
and also Quantum Physics, Relativity Theory and other options in
Theoretical Physics. "I love the beauty and elegance of
mathematics. Even when things get messy you can still end up with a
nice result." Andrew got his BA after three years of study and had not
given much thought about what to do next. |
Andrew
could have stayed on and taken Part III to qualify for a PhD position
at Cambridge, but he did not really fancy doing that. His tutor,
John Knott suggested some places; in particular,
he mentioned Harwell and also the research labs of the Central Electricity
Generation Board (the CEGB was nationalized in the 1990s). "I liked
the research focus of the CEGB and thought they were a very good
organization. Unfortunately, they didn't want me." Andrew did get an
interview at Harwell. "This was quite a serious affair and the
interviews took two days, during which I met a lot of really good
people who did very interesting and exciting things. The place felt
very much like a university, but without the teaching." The procedure
included an interview at Oxford, because Harwell had a practice of
taking people on to do a PhD in Oxford. A PhD position, however, meant
a three-year contract, that would typically be extended for another
three years of postdoctoral research, but not necessarily end in a
permanent position. Andrew was offered the choice of doing a PhD or
starting immediately on a permanent position as Scientific Officer. "I
chose the permanent position for job security." |
|
|
Graduating from Cambridge in June 1975; Andrew is standing in the middle. |
Andrew never worried about not having a PhD. "In
those days, Harwell was one of the best places in the world and top
scientists were working there. I joined the Theoretical Physics
Division, which was a small group of about 30 people. (Incidentally,
the infamous Klaus Fuchs was the first Head of the
Division.)" The Theoretical Physics Division was a very active and
vibrant research environment. "When I worked there, we had three
Fellows of the Royal Society in our Division. The people there were very
brilliant and I found the high-powered environment rather
daunting, but fantastic at the same time. I was very happy with my job
and never looked back or wondered whether I should have done a PhD."
|
|
Andrew joined a
project to develop finite-element codes for computational fluid
dynamics with applications in engineering. "We worked on heat transfer
in fluid flows and this is where I gained my expertise in two-phase
flows and fluid dynamics in general. There were only few commercial
CFD codes around, so we were really doing cutting-edge
research. Harwell was the first place in the UK to get a
supercomputer. I worked with people in the Computer Science Division,
such as Iain Duff and
John Reid. Other well known mathematicians
working there at the time included
Mike Powell, and Alan Curtis. Really, it was an
amazing place to work!" |
Andrew Cliffe with colleagues
John Rae (left) and Juan Matthews (right), from Theoretical Physics
Division at Harwell. |
A Royal Society Industrial Fellowship at
Oxford, 1986-1988
In the 1980s Andrew became interested in bifurcation
theory and chaotic dynamics, mainly due to a seminar presentation at
Harwell by
Brooke Benjamin who spoke on the Taylor
problem. The Taylor problem concerns viscous fluid flow between two
concentric rotating cylinders. At relatively low rotational velocities
the flow is laminar, also called circular Couette flow. However, as
the velocity of the inner cylinder increases, the laminar flow becomes
unstable and other more complicated patterns arise that eventually
develop into turbulent flow with increasing rotational velocity. In
1923, Taylor published his seminal paper
"Stability of a Viscous Liquid Contained between
Two Rotating Cylinders," where he gives a theoretical
derivation of the critical speed at which the instability
arises. "This paper was significant because it was the first time
that anyone had obtained agreement between the theoretical and
experimental stability boundaries. Until Taylor's paper there had
been a crisis in fluid dynamics, because there was no agreement between
stability results obtained from the
Navier-Stokes equations and experiments." Taylor had
chosen his problem carefully: by using long cylinders he was
able to argue that end effects could
be ignored.
Benjamin studied the effect of the ends
of the cylinders by means of arguments from topological degree
theory. "His presentation at Harwell really got me interested in the
problem. I realised that I could do the computations for those
equations and corroborate the theory with numerical results." Andrew
ended up meeting with Benjamin and his postdoc
Tom Mullin. "I visited Benjamin and Mullin in
Oxford. Brooke Benjamin, though extremely eminent, was a really nice
guy. I showed him my initial results and he got very excited about
them. He said I should go back to Harwell and do some more
computations. So I did and I started collaborating with Tom Mullin in
the early 1980s." Andrew clearly enjoyed this project and loves
talking about it. "The 1980s was an exciting time. Chaos and
dynamical systems were new subjects and I was in the fantastic
position of having great computer facilities at Harwell. |
I could
just put a few things together and get results that I am still very
proud of." The collaboration with Tom Mullin was extremely satisfying.
"Benjamin discussed things with us and gave many helpful suggestions,
but he never interfered with our work. He also helped me apply for a
Royal Society Industrial Fellowship." As a result, Andrew got a
visiting position at Oxford for the period 1986-1988 to work with
Benjamin and Mullin. As it happens, this position had an additional
benefit: "As soon as you get an affiliation with Oxford University,
they insist on matching any degree you may have from Cambridge. So,
Oxford awarded me an MA and I ended up having three degrees, but I
only ever really earned one of them!" |
|
|
Andrew Cliffe with his
long-term collaborator Tom Mullin from the University of
Manchester. |
The environment at Harwell in the 1980's was as
close to paradise as Andrew could imagine. "Of course, the UKAEA had a
research strategy. However, if you were able to justify a particular
project, you could more or less do what you wanted. During my
Industrial Fellowship I was supposed to spend half my time at Oxford,
but I ended up being in Oxford almost all the time!" The collaboration
with Mullin continued and later included Mullin's student Simon
Tavener. Andrew also developed links with the
University of
Bath and worked with
Alastair
Spence on bifurcation problems.
Value for money versus adventure and risk
And then times changed... Gradually, there
was not as much freedom to choose research projects and the
pot of money for fundamental research dwindled. "Harwell became a more
commercial organization and I had to work increasingly on more applied
and less adventurous and risky projects. Effectively, I was working
for a science and engineering consultancy company." As a fairly senior
person, Andrew was expected to be more involved in management,
chairing meetings, preparing bids for project, and so on ---
basically, being profitable. "In the early part of my career at
Harwell I had to attend regular promotion panel meetings. This was a
big thing: you had to put a suit and tie on and go to London to be
interviewed by a panel of eminent senior scientists. Their major
concern was whether you published and whether there was evidence of
the scientific impact of your work. For promotion to the career grade
Principal Scientific Officer, management skills became a major
requirement. To be perfectly honest, management didn't interest me
--- this was reflected in the feedback that read his
answers to the management questions were bordering on the
naive!"
In the mid 1990's the UKAEA's nuclear liability was
split off and the rest of the organization was privatised as
AEA
Technology. One area Andrew had specialized in was radioactive
waste disposal. He had worked on a number of projects for what was
then called Nirex, and he had developed an interest in
uncertainty quantification as a result. AEA Technology decided to
concentrate on energy and environment and in 2001 sold off the nuclear
research branch to Serco. Andrew got to work for the part of the
company called Serco Assurance. "The focus of the company was
quite different to that at Harwell when I first joined and I realized
that perhaps I needed a new challenge."
|
|
In 2000 Andrew applied
for an academic position at the
University
of Nottingham. "David Riley was the Head of the School of Mathematical
Sciences and I had known him for a long time. Unfortunately,
my wife became ill at that time and I realized that I couldn't deal
with the additional upheaval associated with changing jobs." Hence,
just before the candidates were invited for interview, Andrew
spoke with David Riley, explained the situation and withdrew his
application. "I felt really bad about this and was convinced they
would probably never consider me for a position again." |
Andrew Cliffe
and the Head of the School of Mathematical Sciences David
Riley. |
Andrew returned to his normal
routine and continued working at Harwell. "I did miss doing risky and
adventurous research, which is not compatible with doing consultancy
work, where the main objective is to complete projects on time and
within budget."
|
Then, in late 2004, Andrew was invited to give a
seminar in Nottingham. After his presentation, David Riley invited him
to his office. "We just had a nice chat and I was pleased that his
busy schedule allowed for it. As I was leaving his office, he gave me
a piece of paper with an advertisement for a position of Chair in
Applied Mathematics." Andrew got home and discussed this opportunity
with his wife. "It was a big deal for us, but my wife was incredibly
supportive. She convinced me I should apply." And this is how Andrew
got offered a Chair at Nottingham. Initially his position was partly
funded by Rolls Royce and he has been heavily involved in the
University Technology Centre at Nottingham. He is working with
colleagues supervising students and postdocs on various projects
related to gas turbine transmission systems.
Academic freedom comes with responsibilities
"My first working day at Nottingham was on the
first of September 2005. I remember the day very well. To get to
work I cycled through Wollaton Park and the weather was absolutely
marvellous." For Andrew it felt like he had regained a long-lost
freedom: finally he could make decisions again on what research to
pursue.
"It was a bit scary too. I was worried about two
things: 1. Could I teach undergraduates? and 2. Would I be
able to get a grant?" As it turned out, Andrew enjoyed teaching from
day one. "My first lecture was to a group of 130 students. When I
walked in, I thought: I am terrified! But it did not last very long. I
learnt very quickly that you never go beyond 10 to the hour, though
--- when your time is up, the students will just leave!" Apart from
that, Andrew is very positive about his students. "The students are
very polite, pleasant and respectful, though some could perhaps work a
bit harder." As to Andrew's second concern, his first three grant
applications failed. "There is a big difference between academic
grants and projects in a commercial context. At Harwell we were in
competition with other companies, but there is not a lot of scope to
differentiate on the science: the main goal is to ensure you are
better value for money. In academia the focus is on the quality of the
science."
|
|
Andrew Cliffe
with Henry Power from the Department of Mechanical Materials and
Manufacturing Engineering. |
|
Overall, it is hard to explain the essential
difference between work at Harwell and at Nottingham. "I didn't leave
Harwell because I was dissatisfied, it was more that I was not very good
at the commercial aspects of the job and this bothered me." Indeed,
when Andrew left, Serco told him he could come back if he wanted do.
"I have a lot of respect for the people who work there and it was not
easy to leave my colleagues and friends." Luckily, Andrew found his
new colleagues to be equally friendly and cooperative. "I really value
my academic collaborations; I've continued to work with old friends
and have had the opportunity to meet and work with new people."
"One thing I really enjoyed about Harwell was that I
never had to worry about whether what I was doing was significant: if
a company was willing to pay for a project then clearly it mattered to
them. I do feel that academic freedom brings with it responsibility to
work on projects that are scientifically significant and important to
the wider community. Much of my current work is concerned with energy
and transport issues and I believe that, in a small way, I am helping
to improve the environment."