We're lucky to be joined today
by Dr. Stephen Morris.
He is the J. Tuzo Wilson Professor
of Geophysics at the University of Toronto
He has a BSC and MSC from
the University of British Columbia
as Ph.D. from Toronto
Where he's been, I think for a while,
10 years? 15 years? I think
Forever
Yes. He's been there for almost forever
and has won two teaching awards
Toronto, also as an visiting fellow
at Cambridge University and is a fellow
of the American Physical Society
At Toronto he directs the
Experimental Non-Linear Physics Group
which, according to its website,
not only does he direct. He is also
the chief fundraiser and receptionist
for that group.
So, he wears many hats.
Thank you for joining us and
taking the time to watch us today.
Hello
I want maybe to start off
if you could talk a little bit about
what drew you to non-linear dynamics or
or chaos in the first place? How did you
get involved in whatever it is you're
involved in?
Well I started out a long time ago
working on liquid crystals,
which are a kind of soft matter topic
a material science soft matter topic
And what's interesting in liquid crystals
are phase transitions, so you
change the temperature and liquid crystal
goes into a different phase
Liquid crystals are very visual. You can
look at them under a microscope or with
your eye and see their phase transitions
nicely, And it's kind of a short step from
there to pattern formation, which is
the normal dynamic systems
where if you drive them with some kind of
force, you increase that force
beyond some threshold and a new pattern
emerges that's what's like a phase
transition. So in fact the
theory of phase transitions and the
theory of bifurcations and non-linear
systems are really close cousins
to each other. And the experiments are
fairly similar to each other
though one of them is an equilibrium
system--thermodynamic system
and the other one is a non-equilibrium
system. But it turns out that is not
so important. Basically when the
experiment is quite similar you can
look at transitions between systems and
patterns just by changing a parameter
and take pictures of it and write theories
It's really quite similar to condensed
matter physics.
Many people of my generation
came here from doing condensed matter
physics, especially soft matter physics
So that's how I got started on it.
I wonder...I mean, my background in grad
school, I was starting with high-energy
theory, which I enjoyed the math of
but it seem to me to be getting
farther and farther removed
from reality and one of the things
that drew to chaos and
non-linear dynamics was it was,
to some extent, the physics of
everyday.
Yes, yes
Sounds like that has been a theme in
your work, as well?
Absolutely. I really like the visual
aspect of it. I like being able to make
experiments out of simple parts
and it's very easy in this field to
just walk down the street and see
something and take in the labe
and make an experiment out of it.
The first one I ever done. It's almost
the exact opposite of LHC
LHC is bigger and bigger and more
and more expensive and takes more
and more people longer and longer
to do the experiment. And in this field
it's most exalted things are
simpler and simpler and cleverer
and cleverer and unique
And they all involve just one person.
Usually, one graduate student gets to
build the first and last experiment
and understand the phenomenon for
the first time.
Right.
And that's really attractive.
And to me, at least, I always felt
that even though they are sort of
everyday and simple phenomena,
the mathematics behind them was no more
or less beautiful than the mathematics
behind string theory--maybe more so
because it actually describes the
real world
Yes
More than string theory.
Yes, I think more so. And it's often very
difficult--one reason these things
weren't done before is--you know--
is the fact that they are difficult.
We still don't understand and lot of
what seems to be everyday phenomena
And the reason is they don't fit the
standard Hamiltonian equlibrium physics
we to teach people. We tell first year
physicists that friction is an annoying
"extra" force that we should ignore
or get rid of in our equations and so on
but it's truth is it is a powerful
organizer of phenomena and you
can't understand fluids without
viscosity and you can't understand
motion without friction
And so, year, it's crucial and
many times these things were looked at
with the tools of 19th Century physics
ages ago, but those tools have been
enormously improved and we have all
kinds of new ideas, a lot of them brought
from condensed matter physics.
And so, on the other side experimentally
you have vast tools we didn't have before,
even digital cameras. The idea now
that you can follow thousand of icicles
or whatever you are studying, and just
images which you couldn't have
afforded to to take that many images
So it opens up--it's like big data
We suddenly have big data on everyday
phenomena we didn't have before.
That means the laboratory experiments can
be very sophisticated because all I'm
using is a commercial SLR camera and
computer, but those things are very
powerful now compared to where they
were 20 years ago.
Sure, yeah. Is there an experiment in
pattern formation one of these one grad
student, one tabletop sort of thing where
maybe you could describe to us a little?
I'm thinking about icicles, because I'm
looking out my window..icicles or
something else?
I have an icicle experiment, this is
one that I'm working on right now
And it's exactly what your think it is
it's a table top thing, it's a big
fridge thing we built on a table top
It produces a small region of minus 10
or minus 20 degrees celsius
You dribble water in the top with a
periscopic pump so we know the rate
of water flow,
You stir the air inside and hopefully
control all the conditions inside for an
icicle. And there is a slot in the side
of the box, you point the camera in there
and you follow the growth of the icicle
over many hours and you can use the
digital image to trace the edge of the
icicle as a function of time, so
you know exactly the way the water runs
out and the humidity and measure the
flow of air in the box. So we can
produce icicles under completely
controlled conditions.
This is almost never done before
a few experiments where people made
icicles and they basically measure their
length and width and they did, you know,
did sort of a gross--how much ice is there
calculation, but the question of what
shape is an icicle is a quite subtle one
This is an example of a moving battery
problem : you have an interface between
ice and air and water and air
and you need to write down differential
equations or models that for the
motions of the interface . So the
outcome of the theory would become
how time distributed the shape of the
growth and length and shape of the
icicle, or any amount of growing for
example. This is incredibly difficult,
no one has ever done it before--
it's much more difficult than you would
think. And there are surprises, so
for example, icicles have ripples on them
Any icicle you might see if you look
out your window have a kind of bumpy
texture. It turns out that those
ripples are weirdly universal.
They're almost exactly one centimeter
in wavelength, everywhere in the world,
always the same. Independent of the
temperature and water flow rate
and everything--no one understands
why those ripple are there
There are also ripples on stalactites,
which are a kind of "rock" icicle
hanging from the top of the cave
and they grow in a similar but not
exactly analogous way. Water runs
over the shape, so the flow of the water
determines the shape and the shape
determines the water flow and for some
reason we still don't understand the
results of the formation of stalactites
So we're also looking for the universal
among patterns. For example, a
rippling pattern that you see is often
driven by quite different underlying
physics, but basically all ripples have
the same symmetry , all have the same
defects, ripples generically move ----
so for example, ripples on icicles
move upwards slowly. This wasn't really
well known. But the experiment is quite
simple, it sits on a desk top and is the
size of a --, and we built it.
Nice. So one of the things-when I think
a little bit about what's different
about chaos or dynamical systems
maybe than other areas of physics
is a little bit--I don't know--maybe a
little about what you are describing
and so if you were trying to come up
with a model for icicle formation
and stalactites, one approach is
to start very detailed through all
basic physics, but is another is maybe
start at a slightly higher level and say,
well, just like you were saying, there
is something common about ripples
so let's see what we can learn about
ripples independent of whether
they are ripples in water or stone
And that, to me, feels it's something
that has a different flavor about
non-linear physics as sort of work you
do and so on. Does that sound right to you?
That's right. Many times one can
qualitatively reproduce universal
behavior using what are called model
equations which are often simply guessed
from symmetry, they are kind of like
the standard model of particle physics
where (inaudible) is guessed by
the symmetry first. And you can
do those kinds of models
and patterns if you know the symmetry
you can write down the basic equation.
And often, that's enough, you don't really
need to give a detailed model of the
underlying physics the model equations
often contain all the dynamics the
patterns have in them from many different
systems, without any more work in a sense,
so that tells you that the features of
that pattern are generic so they're not
related to the underlying physics
they emerged directly from the model
level equations .
This is somewhat analogous
to what people do in phase transitions
or in condensed matter problems
where "coarse-grained" approximations, so
rather than write down the starting
equation for all the atoms and so on,
you write down a simplified Hamiltonian
which contains just a few degrees of
freedom and you hope those are the
ones that are relevant The argument is
always is all the other ones are just
averaged over or disappear or
kind of model, but anyway
it's quite common in condensed matter
physics to make models of that kind
as well. (11:54)