# What is Turbulence?

17 October 2010

## Interview with

Dr Fred Marquis, Imperial College London

## Airplane_vortex_edit.jpg

Ben -    When most people hear the word 'turbulence' they immediately think of being thrown around inside an airplane and perhaps, needing to use those little paper bags that they supply us with.  But the way that it affects flights is just one aspect of a very large and a very complicated subject, as I found out when I spoke to Dr. Fred Marquis from Imperial College London.

Fred -   Turbulence itself is actually quite difficult to define as is.  It's actually a mixture of things and we usually try and classify it.  Some people actually call it a syndrome, but  turbulence is characterised I think by three main things:  disorder, mixing, and vorticity.  By disorder, I mean that if we look at a flow that we call turbulent flow, if we look at a point in that flow and measure very carefully what's going on at that particular point in terms of either temperature or the velocity of the flow, if we look at the data that we get out from that point, we see that it's got very small, sometimes large, seemingly chaotic motions that seem to vary almost randomly as a function of time.  But if we measure for long enough, what we can see is over that long period of time, a mean value appearing as we take our measurements.  So the measurements, if we average them all out, seem to come to a mean or a steady value.

Ben -   What do we mean when we say mixing, when we're talking about turbulence?

Fred -   Okay, probably an easy one for you to visualise is ink mixing into water. If you start off with a nice blob of blue ink and you've obviously got the colourless water.  The blue ink is very easy to see and the blueness of the ink propagates out into the rest of the fluid, and the colour of fluid will become more even over time.  That takes a long time because the mixing that is taking place there is by molecular motion.  If however we take a spoon, we stir up the ink and water, you will see the colour will even out extremely quickly, and that is because as we stir the fluid we're imparting turbulence into the fluid which causes this very high rate of mixing.

Ben -   So, to make something turbulent, do we actually have to add energy?

Fred -   Spot on.  That's exactly what you have to do.  And it's worse than that.  To keep the turbulence going, you have to continually supply that energy because over time, the natural viscosity of the fluid, which you can think of as fluid friction, tends to slow or absorb the turbulent motions, and ultimately convert those into heat.

Ben -   Okay, so that's turbulence and mixing.  The other one you mentioned was vorticity.  Now vortices are these - most people think of them as smoke rings, aren't they?

Fred -   Yes.  That's a good analogy to look at.  A smoke ring is a very well defined vortex.  Again, if you look at a turbulent flow and it could be off the front of a ship or it could be around the bridge support, you will see quite pronounced curly motions.  I mean, even Leonardo DaVinci, in some of his sketching, he was actually sketching these vortical motions within the fluid.  And of course, with our eye, what we tend to see is the big motions, but in actual fact, what's happening is that these big vortices are getting stretched over time.  As they get stretched over time, they get thinner.  Because they're getting thinner, they're getting faster and as these vortices get faster and faster, and longer and longer, they become unstable.  They break up into smaller vortices, so you get a whole cascade of vortices and we call it a cascade of energy.

Ben -   So obviously, turbulence is a very interesting thing, a very difficult thing to study, but why is it important that we do actually research it?

Fred -   Turbulence is important.  The two that I'm perhaps most familiar with are in things like the chemical and pharmaceutical industry, and also the motor industry, in the design of combustion engines and indeed, gas turbines for aero engines.  Now, one of the processes that's going on in combustion is obviously combustion - you're mixing fuel and an oxidant, usually air, together to produce heat and you want that heat to be able to ultimately do work and push your airplane forward or move your car along the motorway.  But how that air and the fuel mix within the combustion chamber determines quite critically the efficiency of the combustion process going on within the combustion chamber.  The efficiency is measured in two ways, if you like.  One way is how much the fuel gets turned into useful heat and the other thing that happens is, during those chemical reactions, pollutants get formed.  The amount of those that you produce is dependent on the levels of turbulence and how efficient your mixing is within the combustion chamber.

Ben -   You also mentioned the pharmaceutical and chemical industries.  In what way do they need to understand turbulence?

Fred -   In exactly the same way, in the chemical and pharmaceutical industry, you've got chemical reactions going on.  Sometimes you've even got competing chemical reactions going on, so for example, if I'm trying to make a drug and I mix two chemicals together, I may well get my drug which is what I want, but I may also get some side products which I don't want.  The amount of side products can be critically dependent upon the levels of mixing that are going on in the pharmaceutical process.  And by controlling the levels of turbulence, for example by controlling how you actually do the mixing in what's called a mixing vessel, you can tune your process to produce the chemical or the drug that you want to produce and try and minimise a side product.

Ben -   So, clearly understanding turbulence is very important, but how do we actually go about researching it?

Fred -   Okay.  If we actually want to try and understand what's going on within turbulence or within a turbulent flow, I suppose you really got sort of two avenues of approach.  The first is an experimental method, so you actually set up your flow, and you carry out measurements.  These measurements are usually laser-based, for example, and we like to use laser-based methods because they don't disturb the flow.  Another way of trying to understand what is going in turbulent flow is that we can actually try to model the turbulence.  When we model this turbulence, we have to resolve all of those chaotic motions that I was talking about earlier.  This means that there's a very high computational cost.  In fact, one of the biggest computers in the world is modelling the weather.  And these weather modelling computers - computer programs, essentially, they have to be large because they're trying to resolve all the turbulent flow within the atmosphere.